Microchip Technology Stock

Good afternoon, everyone. Welcome to day 2 of Citi Global TMT Conference. My name is Atif Malik. I cover U.S. semiconductors and semiconductor equipment stocks. It's my pleasure to welcome Eric Bjornholt, Senior Vice President, Chief Financial Officer at Microchip. I'll kick it off with my fireside questions. If you have a question, save it towards the end, or we'll send the mic to you and you can ask your questions.

Welcome, Eric.

Yes. Thank you for having me. Hi, everybody.

Eric, I'll have to start with the customary, kind of, cycle question. Microchip has now delivered roughly 6 consecutive quarters of above seasonal growth. One of your peers is talking about this above seasonal growth continuing into next year, that's ADI. And your guidance suggests a seventh consecutive quarter of above seasonal growth. So a question is really, where are we in this recovery cycle today? And what are you seeing in terms of traditional drivers like channel restocking and inventories?

Okay. So before I start, I will be making certain forward-looking statements as part of this discussion and refer you to our filings with the SEC that identify important risk factors about the company. So we've been in a period of growth for quite some time. If you look at where we're guiding the September quarter to at the midpoint, it's up 40% year-over-year. Obviously, during that time period, margin profile has changed significantly and doing quite well. But customer order patterns have been very strong. Bookings activity, we mentioned that the June quarter was the strongest bookings quarter that we've seen in about 4 years, and that has continued here into the current quarter, both July and August bookings were good.

So backlog continues to grow. Lead times are staying pretty steady for us. We've seen some pushouts on lead times for some of the outsourced capacity, and that always leads to the question of what are customers doing with inventory. But from what we see, we don't see that there's any significant inventory that's being built at this point in time. Actually, distribution inventory is quite low for Microchip at 25 days. And the order patterns continue to come in strong. sell-through through distribution is strong. And since lead times have stayed very competitive at this point in time, and we are delivering to customers when we say we're going to deliver, we don't think that there's really any significant stocking that has happened yet or restocking at customers.

Great. That's good to know. You talked about the strongest booking quarter in roughly 4 years. Can you help us understand the end markets, were the bookings strong across all end markets? Or there are certain regions or products that are seeing above stronger bookings?

Yes. Bookings have been strong across the board. It's really all product lines, all business units, all geography. There's not really anything to point out as being unique or different or weak in any way, but it's been across the board strength in bookings.

Okay. And how would you characterize the underlying demand trends across automotive and industrial?

So from our perspective, the industrial and automotive recovery was delayed compared to what we've seen in data center and aerospace and defense. But as it stands now, we think that those customers have corrected their inventories from the last cycle, are placing higher levels of orders on us. We produced information last quarter on a quarterly basis for the first time, we used to just do it annually in terms of breaking out end markets. And we showed in that slide that's posted on our website what the growth has been by end market year-over-year. And automotive grew like 29%, something like that. Industrial grew quite nicely. So really, all end markets are growing very good at this point in time, and industrial and automotive are really no different.

Eric, you guys make thousands of parts, microcontrollers, and we noticed that in periods of time when the oil prices started to go higher, the analog stocks, they start kind of stalling because the market thinks that this going to have some effect on your businesses. So in your kind of conversations with your customers, does this topic of oil pricing or this inflationary environment, it comes up? And do you think it has an effect to a certain level on your business?

I think on most customers, it really doesn't have an impact, right? They're looking at their business outlook. And if oil prices impact their business outlook, maybe they pump the brakes a little bit. But we haven't seen that as of yet. Like I said, the bookings activity has been really strong. Design win activity with customers continues to be very robust and some of the new products that we've been talking about publicly, which I'm sure we'll talk about in future questions that you have. So I am not seeing that impact from customers today in terms of what they're giving us in terms of orders and what they're giving us in terms of their outlook over the next year.

Yes. We clearly didn't see it when the Iran war started in terms of the demand profile. I'm curious if you can rank order the regional strength across North America, Europe and Asia for industrial and auto markets?

So we have a little bit less exposure in China, particularly in EV, and you would have seen that when some of our competitors saw that rise in revenue and then the drop in revenue. We really don't have any significant EV exposure in China. North America and European business in automotive is, I would say, trending equally with each other. I don't really see a difference, and we have a little less exposure, particularly in China.

All right. One of the lenses clients are looking at the analog stocks is their ability to supply more in this very strong demand environment. And you mentioned on your last call, you're not at 100% utilization, you have additional tools that can be brought online as demand recovers. But you're also seeing pockets of tightness across the supply chain, particularly OSATs. Where are those constraints showing up most today? And if demand surprises to the upside, can you flex between internal fabs and external partners?

Okay. So first of all, on internal capacity, we are in a very good position. Even though we closed one of our wafer fabs, our Arizona fab last year, we had expanded our cleanroom capacity significantly in the last up cycle. We've invested in tools that you had mentioned before that haven't been deployed yet. So our capital expenditures for the next several years, I expect to be quite low. They have been last year and this year, and I expect that to continue as we grow back into the capacity that we put in place.

We have seen tightness in some of our foundry partner process technology nodes and also with some of the assembly and test vendors, as you've mentioned. I think a lot of that is driven by what's happening in AI and that's squeezing out some of the capacity. So we are doing the best that we can to have forecast from our customers, provide forecast based on history and essentially scratch and claw for the capacity that we need to support our customers and generally have been doing a pretty good job, but we have seen some lead times extend because of what's happening on some of the outside production activities.

Great. If you can just kind of dive into -- you shut down Fab 2, which was 25% of your 8-inch capacity. Remind us how much of your manufacturing in-house versus the external? And then also the back end, how much capacity headroom is left?

So as I mentioned, we shut down Fab 2 last year, which was the smallest of our 3 large fabs. We do not believe that we've limited our growth in any way from that. We do about 35% of our production in-house. The other 65% of the wafer fab is done externally. We took the tools out of Fab 2 that we could use in the Colorado and Oregon fabs, our other 2 large factories. And with that, we think that we can get back to peak revenue plus in our existing fab footprint and are in a very good position from that perspective.

On the assembly and test side, we do more internally. We do about 70% of our production in-house in facilities in Thailand and Philippines, and have good capacity growth capabilities there also from kind of a floor space and land perspective that we had to expand very cost effectively.

Great. The pricing environment has been strong this year. Microchip implemented pricing adjustments effective from mid-August to early September. If you can just talk about what's driving the pricing environment and if you're seeing utilization rates and lead times continue to drive pricing higher?

So we had an issue as part of the last cycle and how that was managed with customer relationships. And that was one of Steve's -- one of his 9 points in his 9-point plan when he came back was rebuilding customer relationships. So we were very cautious about going out and raising prices on customers, although we were seeing supply chain increases over the last 9 to 12 months. So we absorbed those for a period of time. I would say investors didn't really see that because we had cut capacity so much that our gross margins had fallen quite a bit all the way down to 52% last March, and then we're growing quite nicely, but we were absorbing some costs in that.

We feel that our customer relationships are in a very good spot today. And so we decided with some of the ongoing price increases that we were seeing from our supply chain coming through that it was time to increase prices on customers, and we did that in the current quarter, various effective dates but kind of mid-August to early September is when most of that went into place, and that was factored into the guidance that we've given the Street.

That price increase went very, very well from a customer perspective. Many of our competitors were on their second and third price increase already, and it was our first, and it was quite modest. So well received. We are not using this as a way to amplify gross margin percentage. It's a way to maintain that and pass on the costs that we're seeing from the supply chain to customers. And again, they received that price increase very well.

Eric, you and Steve have done a great job in expanding gross margins, but our investors are greedy. They want to see more upside to the gross margins, particularly from the faster-growing data center and aerospace and defense segments. So as you look at your gross margin kind of projection, you're remaining flattish, 66% to 67% through fiscal '27. What are the biggest puts and takes around the gross margins?

Yes. So I want to point out that in the current quarter, when we're guiding to a 66.5% gross margin, we have a couple of onetime benefits that aren't repeatable. One of those is we have an expected higher licensing revenue this quarter. Within that business, we do some sales of intellectual property, and that's happening this quarter. That's the expectation, which will be 100% gross margin business that doesn't repeat after that. And also when we increase pricing, which we did in the middle of this quarter, there is essentially a write-up that has to happen on the product that's sitting in the distribution channel. When we recognize that revenue when we sold it in, it was based on old pricing. And when the quotes were increased, that causes a onetime benefit in revenue increase with no offsetting costs. So that's not repeatable.

But the good thing is, is we expect because we'll have the price increase in place for all of next quarter as well as continuing to ramp our factories and our underutilization charges coming down, that pricing -- excuse me, gross margins can stay relatively flat in the December quarter and in that range of 66% to 67%. So I've been getting lots of questions from investors today is why can't gross margins go higher.

And what I would like to say is we've put out this target of a 65% non-GAAP gross margin and 40% non-GAAP operating margin is kind of what we'd like to achieve through cycle and drift above that in the good times of the cycle and hopefully not drift too far below that when times are more difficult in the cycle. And we believe we'll be able to do that through proper inventory management, which we weren't able to achieve in the last cycle, but I think we're set up well to be able to do that.

So margins are really healthy today. I think you also have to see that we're in competitive markets. Our margins compared to most of our peers are quite high, and we want to make sure that we are pricing our products appropriately in a marketplace where we can gain share and drive operating margin dollar improvement as fast as we can.

Awesome. Let's talk about the data center market. We, the sell-siders, appreciate you guys arekind of breaking out data center revenue, expected to grow 69% to around $1 billion this year, around 17% of sales. So we appreciate all the exercise you went to count all the parts that participate in the data center market. If you can just help us understand why you've seen such a strong inflection and hopefully separate kind of general purpose data center from AI?

Yes. Okay. So we used to break out end markets only once a year. And now we have moved to be able to do that on a quarterly basis, which investors have been asking for. We also used to combine data center and compute together, and we've broken those out separately and compute is only about 3.5% of the overall business.

So with that, we broke the data center into 2 different categories. We have a specific business unit called Data Center Solutions, which purely focuses on products that service the data center market. And that business is growing nicely this year. We'll end this year at about $500 million in revenue for the calendar year. And within that, we've got our PCIe Gen 3, 4, 5, 6 products. We've got our recently introduced retimer products. We've got storage controllers, and we've got memory controllers. So those are pure-play data center products.

We have another section of what we call our standard product portfolio, whether it's microcontrollers, analog, timing, security, et cetera, that sell into the data center market. We also expect that business to be about $500 million in calendar '26. And those 2 combined are going to grow about 69% this year. So we've given that breakout. That's a breakout that you will continue to see on a quarterly basis on an end market perspective, and we're excited about the growth possibilities within that end market for Microchip.

Great. Eric, you mentioned the PCIe product. So we were at the Hot Chips Conference a couple of weeks ago. And we certainly heard a lot of buzz around the PCIe Gen 6 design wins. You guys, I think, spoke about 14 design wins and $100 million opportunity, at least at one program in '27. So just on the PCIe switch and retimer market, how do you see the opportunity? What is your market share right now? If you can just give an assessment on like are there multiple hyperscalers that you're engaged? And where is the momentum coming from?

Okay. So I think you summarized it well. We have 14 design wins on Gen 6 so far. 12 of those are on PCIe and 2 of them are retimer. We've sized that market in 2030 based on external sources at about a $10 billion opportunity on PCIe and about $2 billion on retimers. And we're relatively small players today. Back in Gen 3 and Gen 4, we had a pretty heavy footprint in PCIe. In Gen 5, we missed the market opportunity. We were very late to market. We had done some internal development and were working to develop our own SerDes, and we were just late to market, and we were like 18 to 24 months late. And because of that, we missed out on a lot of the large opportunities when that market was really taking off.

So now we're reentering with a very competitive product in Gen 6. We are the only Gen 6 PCIe product that is on 3-nanometer. That's with TSMC. Customer reaction to that product is very good and what it can provide from a power consumption savings standpoint. And we're excited about the opportunities, and we're targeting hyperscalers, enterprise OEMs, et cetera, with these products, but we haven't really broken out any specific customers. We did mention that one opportunity that you mentioned was about $100 million annually in calendar '27. I should point out that there's really no revenue from PCIe Gen 6 and retimers in the calendar '26 forecast that we provided of $1 billion of total data center as these design wins are really coming to production next year.

All right. Just stepping back and looking at the total data center opportunity, ADI has discussed an opportunity of roughly $1 billion to $1.5 billion of analog content per gigawatt of AI infrastructure deployment. Given your differentiated exposure to timing, power management, connectivity, PCIe fabric, is it fair to frame that your content opportunity per gigawatt is larger than that number?

So we don't look at it in that way. I think maybe our power opportunity is a little bit less than what ADI has just based on the mix of the portfolio that we have. We more tend to tie the growth of kind of our standard products that are going into that data center market more tied to hyperscaler CapEx growth, which is a big number, but we haven't tied it to gigawatts or per blade or anything like this that some of our competitors have.

Okay. Let's talk about the aerospace and defense. Sales doubled from $600 million in 2022 to roughly $1 billion today. What's driving your share gains in the market? And where you think your portfolio is most differentiated?

So we have a very rich history in the aerospace and defense market. A lot of this business came to us through the Microsemi acquisition back in 2018, where they brought us all sorts of products that are radiation tolerant, radiation hardened, have really high reliability and things that make you successful in those markets. And we've expanded on that, continue to invest over the course of the last 8 years. That business was pretty steady through the down cycle. It did experience a little bit of the down cycle, but held a lot steadier. And so it jumped, I think, from fiscal '24 to '25 from like 11% of the business to like 17% of the overall business.

So with all that's happened around the world with the conflict and whatnot, the armaments have been really depleted. And so we think we have a multiyear cycle in front of us as those things get rebuilt. And we are really the largest supplier to the Department of Defense of semiconductor products today. And it's not just U.S., but with NATO's increased investment, the NATO countries' increased investment in defense, I think Microchip is really well positioned for multi-years of growth here.

And I know you guys don't break out within aerospace, what is like satellite or LEOs. But when you look at what SpaceX plans to do, can you kind of expand on your exposure to the satellite and the terrestrial side of things?

So you're right, we don't break it out. What we say is of the aerospace and defense business. Defense is the largest, followed by space, followed by commercial aviation. And they all have various growth drivers behind them. We, historically, with the Microsemi business that I mentioned, have been more kind of deep space versus low orbit but we have products that absolutely can play there.

And because we're known for the reliability and the radiation tolerant and hardened products. I think we are kind of go-to for certain customers in terms of who they want to utilize for these products. So we're well positioned. Some of the lower orbit guys are using more automotive and industrial grade type products, then just putting redundancy in place, but we play well there also. So we've got good exposure that is growing and view that as a nice opportunity for us.

Great. Let me pause here and see if there are any questions in the audience. If you have a question, please raise your hand.

Right. That's good point. Eric, there has been -- have been a kind of a string of M&A activity in the physical AI, edge AI space on Synaptics, then ADI, I think there was some news today. And you guys made an acquisition of Hailo. So my question is really around the strategic rationale behind acquiring a tuck-in type Hailo acquisition? And is Edge AI and physical AI really the next phase of AI deployment?

Yes. So when Steve Sanghi came back to Microchip 18, 21 months ago, we formed a separate product division that we called AI/ML or AI on the Edge. And we have been working on new products in that area. We are excited about this Hailo acquisition. It hasn't closed yet. So we're probably not going to talk in depth about it.

But we actually had some of the Hailo management participate in our annual MASTERs Conference, which we hold in Arizona in the middle of August, where nobody wants to come to Arizona in August, but the engineers do because they get training on the products. And the feedback from customers and distributors was really positive. And we think this accelerates our product road map in that product division by 3 to 5 years. So we're excited about the opportunity. There will be more disclosure on that once the deal closes later this month and definitely on our next earnings call.

Okay. But is this something structurally you're starting to see across with the AI model companies, there's a lot of discussion on wearables using AI. And is it something that you think that the bar is going to be here in 3 years, 5 years? Or are you starting to see some action from your customers?

We are starting to see action from customers today that they're asking for these types of products. And again, Hailo is going to accelerate our product road map pretty significantly. Feedback that we've recently got from customers about the acquisition is very good, and we'll be integrating our teams together. Hailo brings a lot of very talented engineers to us, and we're excited, and there'll be more to come on that in future earnings calls.

Sounds good. Then just broadly speaking on the capital allocation, obviously, the AI data center opportunity is a once in a lifetime kind of an event. Is that changing your kind of capital allocation strategy across dividend and tuck-in acquisitions and all those sort of things?

So it's really not. We've been very focused on getting the balance sheet right after the depths that we went through in the last cycle and leverage has come down significantly. We've started to pay down the debt on the balance sheet, and that is going to continue to be the focus is keeping the dividend flat where it's at for a period of time. I believe that we will end the current quarter with a net debt to EBITDA of under 2.5. So heading in the right direction, but we want to take it lower, and the Board is focused on that, getting it much lower before we start back doing any stock buyback or increasing the dividend.

And we have plenty of capital available to us if these small tuck-in acquisitions continue to come up, whether it's Hailo or something else, and we'd be willing to do that, but you should not view Microchip as doing some mega scale acquisition at this point in time that's more transformative in nature. There'll be tuck-ins to help accelerate product road maps.

Okay. And then the topic around China, it doesn't come up as much, but are you seeing increased competition in China? Is there a threat from like domestic competitors in China? And just comment on your China sales, are they growing kind of in line with the rest of the market?

They really are. I think our China revenue is like 17%, 18% of revenue, and it's been pretty consistent at that level recently. There is some increase in competition, but the bottom line is customers want to design with Microchip because we've got high-quality products, good support, track record of support, having inventory in place for them, and they don't want to change. So we're going to continue to support those customers. Longer term, will competition continue to increase? It probably will. And with that, we're deploying resources throughout the world in areas that we think is going to give us the biggest return on those investments, whether that's in India or other emerging markets.

Great. And then I think you guys -- when Steve puts his focus in mind on fixing things, he does it, and we've seen that. He's delivered on those multipoint programs that he's talked about. From here onwards, where are the priorities for Steve? You guys have got into your gross margin at a healthy level and you're trying to get more supply going and the market is super strong. So what is his focus from here onwards?

Yes. So I mean, we are pretty much done with the 9-point plan. We aren't quite at the operating margin level yet, but we're very close. We'll be there very shortly. So you can almost consider that plan being done. And he is really focused on day-to-day execution, right? Product lines, making sure the investments are being made in the right areas that are going to drive growth for us, making sure we're managing capacity appropriately, whether that's internal capacity, external capacity, focusing on customers and growing the business. So doing the normal things that the CEO does, but the very difficult work that he laid out when he first came back is really completed at this point in time.

Sounds good. Let me see if there are questions in the audience before we wrap it up. One second, mic up.

You talked about the design wins in PCIe retimer. How should we think about that translating into revenue?

So those design wins, there's really nothing happening in the short term. It's more calendar '27 and calendar '28 for all of the PCIe wins that we talked about, both retimer and the switch.

And is there a market share aspiration to size the market to be roughly about $12 billion?

So we have our internal aspirations. It's a large market. We have a long ways to go, and that's not something that we're going to disclose publicly, but we're going to go after every customer that we can and grow the business as quickly as possible.

Is the fact that you were in PCIe 5 an impediment to you gaining share? Or the fact because you were showing PCIe 3 and 4?

So we are well known from a customer base because we've been around in that market for quite some time. This is another business that came to us through Microsemi, and it was their acquisition of PMC-Sierra years ago, very strong in Gen 3 and Gen 4. And it's not that the Gen 5 product is bad. It was just late to the market, right? And that does stick with customers. So we have customer repair and kind of prove it to me that we need to do that we are going to be on time with our products in the future.

And I think our Gen 6 PCIe switch product is proof of that to customers, but we have to say, hey, this is what we're doing with Gen 7, and this is what our goals are from a timing perspective, and then we have to hit those goals. So it's a continuous process. Just because we have a really good product on Gen 6 doesn't mean we're going to win. We have to go out and continue to execute.

Questions? All right. We can wrap it up here. Eric, thank you for coming to the Citi conference.

Thanks for having me. Bye.

Good afternoon, ladies and gentlemen, and welcome to Microchip's Q1 Fiscal Year 2027 Financial Results Conference Call. [Operator Instructions] Following the presentation, we will conduct a question-and-answer session. [Operator Instructions] This call is being recorded on Thursday, August 6, 2026.

I would now like to turn the conference over to Mr. Steve Sanghi, President and CEO. Please go ahead.

Thank you, operator. and good afternoon, everyone. During the course of this conference call, we will be making projections and other forward-looking statements regarding future events or the future financial performance of the company. We wish to caution you that such statements are predictions and that actual events or results may differ materially.

We refer you to our press release of today as well as our recent filings with the SEC that identify important risk factors that may impact Microchip's business and results of operations. In attendance with me today are Eric Bjornholt, Microchip's CFO; and Sajid Daudi, Microchip's Head of Investor Relations.

I will provide a new breakout of our net sales by end markets, including further information on our data on our total data center exposure. Eric will then go over our financial performance. I will then provide an overview of the current business environment and our guidance for second quarter of fiscal year 2027. We will then be available to respond to specific investor and analyst questions.

So let us begin with providing you with a further update on our exposure to the data center market. On June 1, 2026, we provided you a breakdown of our net sales from our Data Center Solutions business unit for calendar year 2025, which we said was $302.7 million. We also said that in addition to these net sales from the data center Solutions business unit, we have many catalog products from various business units that also have exposure to data centers.

In the last couple of months, we have worked to pull together an estimate for this additional exposure to the data center market from these catalog products from various business units. These additional products include our power management analog products mixed signal products, microcontrollers, digital signal controllers, security products, FPGAs, timing products and serially square memory products.

The additional net sales from data centers for these various business units for calendar year 2025 was approximately $288 million. This makes the total net sales from data centers for calendar year 2025 for all Microchip products to be approximately $591 million. And there was the $303 million approximately from Data Center Solutions business unit and $288 million from all other catalog products for a total of $591 million.

This was approximately 14% of our net sales for calendar year 2025. We have also estimated the data center sales growth expected in calendar year 2026. We earlier told you that our net sales from data center solutions business unit is expected to be about $500 million for calendar year 2026.

Now our net sales from the products from all other business units, which also go into the data centers is expected to grow from $288 million in calendar year 2025 to about $500 million in calendar year 2026. This makes the total net sales expected in calendar year 2026 from all of Microchip products, data center solutions business unit as well as all the other catalog products going into data centers as about $1 billion.

So this is expected to be approximately up approximately 69% from $591 million net sales from data centers in calendar year 2025. So a growth of about 69% from $591 million net sales in calendar year 2025 to about $1 billion in calendar year 2026. Now a little bit about calendar Q1 and Q2.

In calendar Q1 2026, which was the March quarter, our net sales from data centers was up 77.2% from calendar quarter 1, 2025. In calendar Q2, 2026, the June quarter just ended, our net sales from data centers was up 97.8% and from calendar Q2, 2025. As our numerous new design wins on our PCI Gen 6 switch, PCI Gen 6 retimer, storage controllers, NVM controllers, power management products, mixed signal products, security products, timing products and memory products proceed to production in calendar year 2027, we expect significant growth from data centers in 2027 and thereafter.

Now here is our end market net sales breakdown for the June quarter. We remind you that these percentages are our best estimates of the end market splits. However, there is probably a couple of percent error band due to the fact that about 50% of our business and the long tail of customers are serviced through distribution, which makes it difficult to track the end market.

So our June quarter end market breakdown was as follows: Industrial was 32.2% and Data Center, 17.1%, Aerospace and Defense, 16.7%, Automotive, 15% and Communication, 8.2%; Consumer Appliances, 7.4% and Compute 3.4% and you can see that we're now breaking out end markets of data center and compute separately. So we are now breaking out our sales into 7 end markets.

From June quarter 2025 to June quarter 2026 saw year-over-year growth, our industrial net sales grew 24.3% and data center sales grew 97.8%. Aerospace and Defense grew 45.6%, Automotive grew 29.3% and Communication grew 53.3%. Consumer appliances grew 19.1% and compute grew 9.6%. So this is the most comprehensive breakout you have heard from us, you have to unpack for you. With that, I'll pass it on to Eric.

All right. Thanks, Steve, and good afternoon, everyone. We are including information in our press release and this conference call on various GAAP and non-GAAP measures. We have posted a full GAAP to non-GAAP reconciliation on the Investor Relations page of our website at www.microchip.com, and included reconciliation information in our earnings press release, which we believe you will find useful when comparing GAAP and non-GAAP results.

We have also posted a summary of our outstanding debt and leverage metrics on our website. I will now go over some of the operating results, including net sales, gross margin and operating expenses. Other than net sales, I will be referring to these results on a non-GAAP basis which is based on expenses prior to the effects of our acquisition activities, share-based compensation and certain other adjustments as described in our earnings press release and the reconciliations on our website.

Our non-GAAP financial results were all above our guidance provided on May 7, 2026. Net sales in the June quarter were $1.485 billion, which was up 13.2% sequentially and up 38% from the June 2025 quarter. We have posted a summary of our net sales by product line and geography on our website for your reference. On a non-GAAP basis, gross margins were 63.8%, including capacity underutilization charges of $38.5 million.

Operating expenses were at 28.7% of sales and operating income was 35.1% of sales. Non-GAAP net income was $438.6 million and non-GAAP earnings per diluted share were $0.76, which was $0.07 above the midpoint of our guidance. On a GAAP basis in the June quarter, gross margins were 63.2%. Total operating expenses were $602.1 million and included acquisition and intangible amortization of $90 million special charges of $18.9 million, which were primarily driven by 2 long-standing legal matters, which were settled during the quarter as well as our activities associated with the closure of Fab 2 share-based compensation of $66.3 million and $0.4 million of other expenses.

The GAAP net income attributable to common shareholders was $202 million or $0.37 per share. Our non-GAAP cash tax rate was 7.5% in the June quarter, and we expect to have a tax rate of about 7.5% for all of fiscal year 2027. Our tax rate is down from the prior year for a variety of reasons, including the amortization of domestic research and development, experimentation expenditures capitalized in previous years, and the impact of a reduction in inventory reserves.

Our inventory balance at June 30, 2026 was $1.047 billion. We had 175 days of inventory at the end of the June quarter, which was down 10 days from the end of the March quarter. Included in our June ending inventory was 14 days of a long life cycle, high-margin products whose manufacturing capacity has been end-of-life by our supply chain partners. Inventory at our distributors in the June quarter was at 25 days, which was down 1 day from the March quarter and at the lower end of what we have experienced historically.

Our cash flow from operating activities was $511.5 million in the June quarter. Our adjusted free cash flow was $478.6 million in the June quarter. As of June 30, our consolidated cash and total investment position was $272.3 million. Our total debt decreased by $138 million in the June quarter our net debt decreased by $170 million. Our adjusted EBITDA in the June quarter was $587.8 million and 39.6% of net sales -- our trailing 12-month adjusted EBITDA was $1.798 billion.

Our net debt to adjusted EBITDA was $2.85 at June 30, 2026. We expect our net debt to adjusted EBITDA to reduce significantly as we progress through fiscal year 2027. Capital expenditures were $13.9 million in the June quarter, and we expect capital expenditures for fiscal year '27 to be about $100 million. Depreciation expense in the June quarter was $37.8 million.

I will now turn it back to Steve, who will provide some additional commentary on our June quarter results and our guidance for the September quarter. Steve?

Thank you, Eric. I will now reflect a bit on our performance in the June quarter. The June quarter was an excellent quarter for Microchip across the board. We beat the high end of our guidance in net sales and our non-GAAP financial metrics, like gross margin percentage, operating expense percentage, operating profit percentage and earnings per share all beat the high end of our guidance.

In all of these metrics, we continue to make excellent progress towards the long-range targets. Our non-GAAP gross margin percentage, we are at 63.8%, which was a sequential improvement of 222 basis points. Our non-GAAP operating margin was 35.1% in the quarter, which was a sequential improvement of 452 basis points. Our inventory and days of sales came down again last quarter. We're getting pretty close to a range of 130 to 150 days of inventory.

We will now take the focus off of inventory and instead focused on handling the large growth we are experiencing and deal with various constraints in foundries, OSAT and internal production. I will now talk about our business environment. We believe that we have completed the distribution inventory correction.

Our overall distribution inventory is now low and some replenishment needs to take place in the coming quarters. We saw a small decline in distribution days of inventory in the June quarter. Our distribution sell-through grew by 17% sequentially in the June quarter. This is a result of distributors' customers completing their inventory correction and starting to reengage in buying our products.

Our customer count is now going up. We are also seeing customers from new designs and from our improved relationships start buying our products adding to our bookings, revenue and customer count. Now let's get into guidance for the September quarter. Our bookings for the June quarter were very strong, the book-to-bill ratio for the June quarter was well above 1, resulting in a higher backlog entering the September quarter compared to when we entered the June quarter.

The June quarter was also the strongest booking quarter in about 4 years. A comment about lead times. Lead times for a majority of our standard products have been 4 to 8 weeks for some time. But now with a dye inventory and finished goods substantially reduced, we are seeing lead times stretch on many of our products. We are running into challenges on certain kind of substrates, subcontracting capacity and foundry constraints on multiple nodes.

These challenges previously were isolated to specific areas but have now spread broadly to many of our products that come from foundry. The challenges in subcontracted capacity and test capacity are also stressing our lead times from products built by our own fabs. Our customer request for expedited shipments have increased significantly from a few quarters ago. We are receiving a large number of orders for requests inside the lead time and many times are going unsupported in the quarter.

We are advising our customers to give us longer-term backlog to match at least our lead times and preferably matching our manufacturing cycle times. I also wanted to give you an update on the pricing. The price adjustments we discussed in the early June have now been successfully implemented in the marketplace most pricing is effective between mid-August to early-September, making a small impact to the September quarter.

Taking all of these factors into account, we expect our net sales for the September quarter to be up 8% sequentially, plus or minus 1%. This at the midpoint will be up 40.6% from the year ago quarter, we expect our non-GAAP gross margin to be between 66% to 67% of sales. We expect our non-GAAP operating expenses to be about 27.5% of sales. We expect our non-GAAP operating profit to be between 38.5% and 39.5% of sales. We expect our non-GAAP diluted earnings per share to be between $0.91 and $0.95 per share.

At the midpoint, non-GAAP earnings per share would be up 165.7% over the year ago quarter and up 22.4% sequentially. Our guidance on non-GAAP gross margin percentage obviously raised some questions because it exceeds our long-term target of 65%. There are a lot of factors impacting this guidance First, there is some excellent product mix in there. Then there is our 100% gross margin on an expected very strong quarter for licensing business unit. Then there is price increase capturing the inflationary costs that we absorbed in the prior quarters. Then there is lower new inventory write-off versus prior quarters.

And finally, there is lower underutilization charges as we are ramping up the factories. So as you can see, there is no single item for this gross margin, there are a lot of items that we have been improving to get to our financial model. The net result is that we are very close to our financial model with gross margin percentage above it and the operating margin percentage just a shade below the model.

At this point, we will leave the long-term model unchanged. You should not expect non-GAAP gross margin percentage to continue to rise above this level. With that, operator, will you please poll for questions?

[Operator Instructions] Our first question comes from the line of Timothy Arcuri from UBS.

Steve, I wanted to ask you just your perspective on how long this upturn can last. So September is going to be the sixth quarter above seasonal. I know it's a unique time, and you're coming off a big downturn. And it does sound like you think both the disties and the end customers are just beginning to rebuild inventory. But usually, you're not about seasonal for more than 8 quarters. You know that better than anybody does. So how do you assess this? How do you think about how long this can go on?

There are at least 3 or 4 factors going on that may make this last a bit longer than usual. The first is the Data Center. Data Center exposure is going through significant growth. And we just talked about growing almost 69% this year. So I think that's 1 that could last longer. The second 1 is Aerospace and Defense. The Aerospace and Defense buildup is just kind of just beginning. We're getting lots of orders for missiles and drones and radar installations and interceptors and barrel tanks and ammunition and all kind. And that's a multiyear buildup.

So I think that could last for a while. The third 1 is while you say that it's been above seasonal for 8 quarters, I haven't countered. I don't know how many exactly, but I think the 2 markets, industrial and automotive started much later. They've probably been just recovering maybe only about a couple of quarters. So I think there is a lot more to go on the industrial market as it is growing a lot of the AI is going into industrial and industrial is also a very broad market, our largest market segment that could go longer and other is the automotive which is probably still not extremely strong. I think automotive market is just kind of beginning to recover. So I think those are the factors which could make this upturn last a bit longer.

And then Eric, just super fast. So are there any inventory sale of previously written down inventory that is inflating margins?

So I would say we're kind of like we said last quarter, that inventory reserves have kind of normalized at this point in time. We did reserve quite a bit of product based on our accounting policies and where the business was at during the down cycle, and there is sell-through, there's sell-through every quarter.

And new reserves are coming down. So it definitely is a net benefit margins, but we don't think that's going away, at least in the short term, with where we're at.

Your next question comes from the line of Vivek Arya from Bank of America.

Steve, I wanted to talk about the data center -- what is the right way to model your data center growth for the next 2 to 3 years? Just conceptually, what are the moving pieces? And how are you benchmarking whether you are gaining share, losing share, keeping share? Like what is the served addressable market for you? How fast is that growing? And how do you expect Microchip to do relative to that market?

Vivek, our data center exposure is probably the broadest of any other company. Many companies could have exposure, let's say, on PCIe switch or could be just in a power management going to the act. As I read through my summary, I mean it's just -- it's very, very broad. It's power management, it's memory, security products, microcontrollers analog parts, mixed-signal parts, switches, controllers and others.

And it's difficult to assess what is the total TAM in each of those markets in data center and how much our competitors have and how much we have and really try to assess all that. We have done a lot of work on this one. on the request of investors and analysts, and this is probably the most comprehensive breakdown we have given you, and we'll continue to give you update, but I don't really not able to project it over the next 2 to 3 years and answer your question, are we gaining share or losing share? I would like to think we're gaining share.

And maybe if I could follow-up on gross margins at 66.5%. I think you mentioned we should not expect more than that. But if you're putting in pricing just in this quarter, you're not really reducing inventory anymore. Mix seems to be getting better. So why shouldn't there be more upside to gross margin? If there's no more upside as this kind of level sustainable. So just kind of the puts and takes on gross margins right beyond this very impressive 66.5% midpoint that you're guiding to?

Yes. So it's a good question, and we knew we'd get it. So Steve provided some commentary on the various things that are impacting it. But there's a couple of things that are not really repeatable that are happening this quarter. We do expect a larger licensing revenue, which is 100% gross margin this quarter. That business tends to be lumpy, and this will be a lumpy quarter from licensing. When we do a price change, and I think you've heard others talk about this in distribution, that impacts the inventory that is sitting in the distribution channel and that's kind of a onetime impact in terms of the revenue that will be recognized or expected to be revenue recognized from that.

So there's a lot of moving parts, obviously, we do still have things working in our favor with improving utilization and whatnot. But it's a balance and 66.5% is a very healthy margin is how I describe it.

What are those one-offs, Eric? Sorry, how much?

We haven't broken those out, Vivek, but they are definitely impacting the margins favorably this quarter.

So we think the entire quarter of increased prices, although there are some expenses, too. We're getting cost increases also from foundries and housing all that. So there's a balancing on that too. But full quarter of increased prices will balance some of these onetime costs that you talked about or onetime benefits you talked about.

One is a very rich mix, product mix in licensing and second being this onetime benefit from the inventory and distribution. Those 2 will be negative for the following quarter, but the full quarter price increase will be positive, and it will probably balance and gross margin could be still in that range, but don't expect it to go higher.

Your next question comes from the line of Joe Quattrone from Wells Fargo.

Yes. Maybe just curious on the data center business. How do you think -- when you look at your design win pipeline, as you think about 2027, how should we think about just the potential for accelerating growth?

Well, like I said, it's -- we have very, very broad exposure. If you were to start to look at design wins across our microcontrollers, memory products, power management and others -- and 1 of the reasons it took us that long and it's more difficult to do is on data center solutions business unit, which is like the Gen 6 switch and retimers and others, there, we engage with very large customers.

We engage with hyperscalers and enterprise customers and winning large design wins. But when you look at our catalog product broad-based, there, we don't win designs that hyperscale as an enterprise. We win designs at power supply companies and card stuffers and module makers and companies like Leighton and Delta and others, which then are selling their products into data center business unit.

So across our entire portfolio, there will be hundreds and hundreds of designs. And it's very difficult, therefore, to answer that question. We will limit the design win information to really the data center solutions business unit. And there, we could tell you that last quarter when we came to you, we had 6 design wins. And at the end of the quarter, we had -- this quarter, we had 12 design wins. And as of today, we have 14 design wins, 12 on the Gen 6 switch and 2 on the Gen 6 retimer.

And many more in the works. So I think that 1 is easy to identify. But if I were to start to identify design wins, the analog catalog products or memory products or security products, they're in less.

That's helpful color. Then maybe just -- you talked about things increasingly getting tighter from a supply perspective. I know you guys have maybe some tools that are still yet to be installed or put into production. Are we getting close to where that might be a possibility? Or is it more the tightness, particularly on the back-end side.

We don't have shortages on our internal fab capacity. We're still not at 100% utilization on our tools -- and like you said, there are many more tools available that could be readily installed. We have not yet started installing them because we still are not at a full utilization, but many of internal products get fimbeled and tested at the subcontractors on advanced packages and different test platforms, and that's where we have some of the constraints for the internal fab capacity.

And on the foundry, we have constraints on many nodes -- and if some of the foundry wafers are coming inside for assembly and test, they were doing better if they're going outside for assembly and test because there on some advanced packages and test platforms, then they are getting squeezed by AI capacity that's crowding out.

So it just depends on what the combination is. If it's inside assembly, inside test and inside fab, that is the best combination. Anytime there is foundry involved or outside OSAT involved, then that's where some of the constraints are.

Your next question comes from the line of Christopher Rolland from Susquehanna.

Sorry, I think there was a bad connection there. Yes. So my question is just about serviceability of upside. Both external and internal, if this cycle will be done like what is the game plan here? And would we ever consider a 300-millimeter fab? Would that ever come back into the thinking here?

Yes. So when we talk about constraints, you shouldn't interpret that as there is no additional capacity available, and there's no additional growth available. Constraint simply means we could use more than what we were given if you've given more the upside could have been a little larger. But every quarter, we have incremental capacity, fabs and assemblies and tests are growing their capacity, and we have increasing allocation on many other products, we were constrained in March quarter 2 yet in June quarter, we were able to post substantial growth, and that's what we're planning to do for the September quarter.

So yes, there is more assembly capacity available. Yes, there is more test capacity available. We're also bringing some products from outside to inside and growing our insight assembly and test capacity -- and we're getting increasing allocation on a lot of the constrained nodes from this quarter to next quarter. So you should not, by any reason, interpret a world constraint to be that the revenue is capped here, not at all.

I think the second part of your question was the possibility of 300-millimeter. And that's not in our plans today. We will rely on our foundry partners to provide anything that we need on 12-inch and the technologies that it supports.

We have hundreds and hundreds of masks running on 300-millimeter and trying to put our own 300-millimeter fab in technology, which will not be compatible with the foundry technologies because we'll have to build our own, then all those parts will have to be redesigned. It's just not a practical thing to do. And so there's no plan to have around 300-millimeter fab.

And perhaps as a follow-up, just because AI has just been such a growth engine and you have PCIe, some other AI products here as well. just going to be an area of focus that you guys may be looking to bolster either organically from here or even inorganically moving forward?

So we're not looking for any large acquisition inorganically. We recently announced a small acquisition inorganically, which is a company called a Halo which is headquartered in Israel, and that deal should close sometime in September. It's revenue is small, but its products are outstanding.

Company ran into some financial troubles, and they had to restructure and became available and we bought it. But the benefit it provides to our road map is just totally incredible in the area of AI on the edge, it advances our road maps by about 5 years. So the products they already come to us -- it was going to take us 4 to 5 years, about 4 years to produce those products in another year to market them and get the design wins and all that, which they already have.

So significant benefit from that acquisition but it's not a very large acquisition that moves the needle. I think it will move the needle on our clock because its revenue should grow pretty rapidly because it's basically primed and all the resources of Microchip selling it, we could grow it rapidly. But in the very first quarter, it's a small revenue.

Your next question comes from the line of Chris Caso from Wolfe Research.

For my first question, perhaps you could comment on December quarter seasonality. I think typically, it's down a bit for you. It looks like consensus is up a little bit in the December quarter as it stands now. But I guess the question is with the different mix of your business, how does that affect seasonality going forward? And what are the factors that might cause this December to be different than whatever would be typical.

So I think what we are willing to say is usually our December quarter is down 3% to 5%. And we think the quarter would be better than seasonal. We're not willing to comment on what the long-term new seasonality would be as some of these markets were in Data Centers, in Aerospace and Defense and others become more mature, I think a couple of years out, maybe a different seasonality.

But for now, I don't know the answer to the long term, but short term, the December quarter should be better than season.

As a follow-up, it does appear that we're in the recovery part of the cycle now. Steve, I know that you've been a little cautious about cash flow and getting that level down and putting the company in the right place with an apparent recovery with happening now, some better cash flow going on, maybe you could update us on your view of cash return and what to do with cash that you're going to be generating?

So we still have substantial debt, about 5.5% is...

Yes, that's about $5.2 billion.

Yes. So net debt is about $5.2 billion, growth is a little higher than the closer to 5.5%, maybe even though the leverage ratio has now come below 3, we believe the debt level is still too high, and we will continue to use the entire cash that's available beyond the current dividend and use it to pay down the debt for a foreseeable future. There's just no other plan. No plan to buy back stock and no plan to increase the dividend for the foreseeable future.

We're very pleased about the progress we made last quarter on our reduction in the net debt to EBITDA dropped to 2.85. I would with our guidance, it will drop below 2.5% this quarter. So making good progress.

Making very good progress and hopefully heading to a number where it starts to drive different actions, but we're not there yet.

Is there a particular number you have in mind that would drive different action?

Not that I'm willing to share.

Understood.

Your next question comes from the line of Vijay Rakesh from Mizuho.

Steve, just going back on the data center side. Is there a way to look at what your content is like dollar per kilowatt per rack? And in terms of the 12 design wins, any of them on the scale up side on the rack, where do you see that -- those wins if you can give some more color and a follow-up.

Trying to understand the question is?

No, we do not give any sort of direction in terms of dollar content per rack. So that was the first question. Of the 12 design wins, it is my understanding that we have both scale up and scale out opportunities within that. Sajid, do you confirm that?

It's mostly at the core level -- at the vendor level. So. Yes.

Basically', I switch connects the GPUs to CPUs and CPUs to memory and it's used if you scale out or scale up in both cases, and that's where all our design wins are. There are some hyperscalers in there. There are some enterprise customers in there. It's a pretty broad mix in those 14 design wins.

Got it. And Eric, on the gross margin side, is the assumption that we should expect it to be kind of in this flattish going forward?

Yes. I mean, what we said earlier is that we -- you shouldn't model in that it's going above the 66.5%, I'll call it, that we're guiding to at the midpoint of guidance. It's really strong gross margin that's above the long-term model. We're continuing to make improvements, but there's definitely some headwinds on the cost side that we're absorbing also. So we're comfortable that we're going to be in this -- about this range for at least the next few quarters.

Your next question comes from the line of Harlan Sur from JPMorgan.

If I look at the mix of your direct versus your distribution customers, the June quarter marked the second consecutive quarter where your direct customers only grew like mid- to high single digits sequentially whereas your Disti business grew strong double digit sequentially in March and June, right, and on a year-over-year basis for both March and June is Disti outgrew direct by around 2x. I know it's harder to track your direct customers internal inventories, but are they still maybe working through some excess inventories? And does that imply direct customers are a potential future tailwind for growth when their inventories start to normalize?

I think I would look at it more of where the distribution network was coming from with very elevated inventory when you look back a year ago, and working through that and now getting down to what is a very reasonable inventory level at the 25 days that we quoted.

And so then kind of returning to a normalized level where now they are needing to purchase kind of in line with what the consumption is. And they've got a lot of customers that are coming back that were over inventory that now they need to support. So I view it more from that perspective. I don't know if Steve has any different view than that.

I think when the products were constrained, the direct customers got better supply than the distributors did -- and when the market essentially came down and the supply became available, that's when the distribution got served. So distribution built inventory later -- and that's why it took longer for distributors to correct. And as they corrected now you're seeing the benefit. I think the OEM customers were slightly ahead of it.

I appreciate that. And then, Steve, within your Aerospace and Defense business, this segment underperformed our overall growth profile last fiscal year. this segment is starting to drive some pretty outsized growth where it was up 46% year-over-year, 20% sequentially in the June quarter, as you highlighted, defense spending by the U.S. and global government.

You've got global commercial space programs are all quite strong. You guys have a very strong leadership position in RadHard FPGAs, RadHard, MCUs, interface, timing, power. Like do you anticipate growth outperformance for A&D for fiscal '27, but more importantly, does the multiyear growth profile look like continued outperformance given some of these programs have very long program lives?

Qualitatively, yes, but quantitatively, don't dial in at 46% growth because we're talking about these growth numbers of some pretty depressed levels in the March quarter 2025, our total company revenue was $970 million, and we're now guiding to 16.03% at the midpoint. So there has been a massive growth from the bottom in March of 2025. So when I do year-over-year comparisons, they obviously look very, very good, which is great, but just don't dial those kind of numbers going forward.

But qualitatively, yes, we are very positive on growth on Data Center. We are positive on growth on A&D. And we think the Industrial and Automotive are trying to catch up there were later recoveries in industrial is probably a little ahead of Automotive, and that makes up all of our major segments. The rest are small.

Communication is small, Consumer is small Compute is small. The 4 large segments are Industrial, Data Center, A&D and then Automotive. 2 out of those Data Center and A&D are very strong, and Industrial and Automotive are in catch up.

Your next question comes from the line of Blayne Curtis from Jefferies.

I had 2. I just want to go back to the conversation about normal seasonal and if you're shipping above. Is there any way to kind of think about what pricing was as a tailwind in September? I think your seasonality is usually maybe up 4%, 5% year-to-date. So just kind of trying to understand, I think you said it was done by September 1. But can you quantify what the impact was?

The impact of pricing on September quarter was fairly small, basically going to get less than a month out of the 3 months. A lot of the customers pulled some orders ahead to into they pulled some orders prior to the higher price becoming effective. Within the quarter, they pulled in from September into August before the price increase on August into July before the price increase.

So stayed within the quarter, but many people pull the orders in, trying to really avoid the price increase for a few more days, a few more weeks. So the impact on September quarter is fairly small, but the impact on December quarter will be the full quarter.

And then I want to ask you, is this kind of a onetime thing that you're going to give us these end markets? Or are you going to break this out more regularly on a quarterly basis versus annual? And then within that, I guess, the only other segment that was up nicely was communications. Maybe you can talk about what's going on there?

So our intent is to provide it more regularly. And in terms of communication we are seeing a significant recovery in communication markets from 5G, from other communication racks, the major communication customers, a lot of them in Europe, we're seeing significant recovery in that segment. It was pretty good growth. Now that sector was quite depressed also a year ago, when you look at year-over-year numbers, the year ago numbers were very low.

So a lot of that good number is because the prior year numbers were very low, although we're seeing recovery too. It's a smaller segment. It's about 8% of our business. So it's not as pronounced as the larger segment, but communication, we have seen recovery.

So I'm just going to be a little bit more clear in Steve's response on the end market breakout. It is our intention on a go-forward basis to break it out quarterly. That's what analysts and investors are asking for. We've set up a process to be able to do that. So we'll share it with you at the end of each quarter.

your next question comes from the line of Tom O'Malley from Barclays. .

You mentioned margins kind of staying at these levels on a go-forward basis, but we've been hearing from others in the space, just a strengthening of input costs that are causing some margin headwinds. So could you maybe call out the areas in which you're seeing the most pressure? And then how are you able to offset those?

You guys mentioned 1 of your line items on gross margins is obviously offsetting some of that with pricing increases. But do you feel like you have the flexibility to continue to do that on a go-forward basis, if the cost environment keeps getting worse?

We did this price increase with the intention of a onetime price increase in which we capture some of the costs we have been absorbing in the prior 6 months. So we increased the price on those parts also. We didn't set it up where we're going to do it quarterly or every 6 months price increase. There has to be another event, a major price increase by the foundries and OSATs on us.

And if that happens in 2027, then we'll take a look at it again. But currently, there is no expectation we've given to the customers that this will be a routine price increase every so often. This is basically for now, it's 1 and done. And a new price increase will be driven by a new set of events.

Yes. And we are continuing to ramp our factories, which will provide us better utilization decreasing on the utilization charges, which hopefully can offset some additional cost increases that are likely to come in.

Some of the place I don't already have knowledge of some of the price increases where we had already knowledge of which haven't kicked in, they're kicking in either in the fourth quarter on January 1. We tried to dial in into this price increase.

Yes. Just as a follow-up, I don't want to get greedy here because you guys are providing a lot more color and we do appreciate it. But in terms of the growth rates of these different segments, you tried your best earlier to kind of capture data center. But some of this other stuff you now have multiple years of data that I'm sure you can see what it's been growing at.

Any color on what you expect each of these segments to grow longer term? Are you going to save that for an Analyst Day or something down the road? Appreciate it.

I think they're very widely. I mean just these things very widely. They're very lumpy and 1 year, Industrial could grow other year Automotive could grow, they go through different cycles of inventory correction communication could go based on 4G going to 5G, there's a 6G. I don't really think there is anything like you can say, CAGR for this segment is that going forward, we don't think there's a number we can put together.

Like you said, we have really put a lot of effort in providing this breakdown. So be nice, so we keep going here.

Your next question comes from the line of James Schneider from Goldman Sachs.

I was wondering if you could maybe just kind of comment on the behavior you're expecting to see from distributors. You talked to -- talked about 25 days being at the low end of the historical range. when you talk to what's your sense about the levels they want to get back to what's realistic to expect over the next, say, 2 to 4 quarters in terms of a healthy channel inventory. Do you think they want to get sort of back to the normal levels or potentially even above that to build on the stock?

So I personally met with some of our largest distributors only in the last 3 weeks. And what the distributors are seeing is that -- they know their inventory is low, and they would like to get the inventory to a normal level, but they would like their customers to place the orders on them so they know what to place the order on us.

And they're seeing, despite the repeated warnings to their customers, and all the talk that's happening in the industry where AI is crowding out everything. There are plenty of articles on longer lead time. There are various analyst reports on lead time growing. Companies are talking about it, the distributors have frustrated that their customers are in listening and aren't giving them longer-term orders, which is -- that's what has happened in the past.

Customers don't move until they get in trouble. They have to go some lines down and then the panic. And I think in each cycle there are a new set of purchasing managers, they move around the old ones retire, the new ones come in, in every cycle. And I can say the same thing about our direct customers. Some of them have to go lines down and then they ask for expedite and we say, we can't do it or it will cost money to do it, and that's when they wake up and they give you longer-term orders.

So distributors are not getting orders from their customers and they're complaining and some of them are planning to give us some orders based on their expectation of longer-running designs where they think they will get the order or they know what to order. But they're not able to order across the board because they don't know what to order.

Yes. I guess some things never change. And a follow-up. Steve, if you provide some good commentary before about your internal capacity, both on the front end and the back end, realizing you're continuing to add capacity on margin, and you had mothballed a bunch of capacity in the downturn. Can you maybe give us any kind of sense as you think about kind of re-expanding that capacity from the current levels of internal capacity you have today, if you -- without doing anything unnatural how much more capacity could you support above that level? Is it 10%, 20%, 30%, et cetera?

It's higher than the numbers you're talking about. We got substantial capacity which is unutilized. And if you keep ramping to use all the installed equipment, we could grow significantly. And then there is $450 million of equipment that's still sitting back, some of that not even installed and that can take the capacity a lot higher.

So rather than trying to put the numbers on it, the internal capacity, fab capacity is not a constraint. In most cases, internal assembly and test isn't constrained either because we could add more capacity, we're adding as we speak, where the constraints are is 1 in the foundry and 1 in the -- the second 1 in the OSET. There are a lot of packages we don't do internally. We don't do PGAs internally.

We don't do many of these packages that require substrates and advanced packaging and multichip and other, we do outside. And AI is crowding out a lot of that capacity, and that's where we're getting constrained. Now our major suppliers are building more factories and ASCs building more factories and Amcor is building more factories and others are building more fractories. So the capacity is growing and we're getting our share of it.

And that's what I said, don't think of constraint as there's no growth available, but we're not getting everything we need. We can make higher growth if we had everything.

Your next question comes from the line of Quinn Bolton from Needham & Company.

I guess Steve, Eric, I wanted to just sort of follow up on the unutilization charges. I think you said $38 million this quarter. It's about 260 basis points -- you didn't mention it as 1 of the margin tailwinds that was driving the gross margin to 66% to 67% next quarter. And so -- how should we think about those underutilization surges kind of decreasing over time?

Are you kind of keeping those and they can largely offset input price increases beyond the September quarter to help you kind of keep margins in that 66% to 67% range? Or is there a different way we should be thinking about the roll off of those underutilization charges?

You're thinking the right way. What we are asking you guys is to take that underutilization charge as it goes away and not added to the gross margin because -- like we said, we have we got onetime gross margin benefit in this quarter from distribution inventory on which we have taken a price reserve.

And when the prices are going up, that reserve was down -- so that becomes a onetime 100% gross margin. And then there's a very, very good licensing quarter. It's lumpy. And this quarter, it's very good as it goes back to normal, that 100% margin goes away. So there are lots of moving parts. And as those onetime things roll off, the utilization is improving and the number you talked about then helps to keep the gross margin in that range rather than take it higher.

Yes. And the underutilization improving was mentioned by Steve in his prepared remarks about being a help in the 66.5% midpoint guidance this quarter. And just on the underutilization charges, last quarter, I think they reduced by about $8 million sequentially. I'd expect about the same in the current quarter.

He was taking the remaining underutilization charge, dividing over the revenue coming up with 238 basis points improvement in seeing where the gross margin doesn't go up. And what we're seeing is that there are a lot of onetime things and it will offset those, yes.

Got it. No, I think that's pretty clear. And I just wanted to ask on auto. Most of your peers year-on-year probably grow in their auto business 15 or so percent, I think you're up nearly 30%. So just -- do you think that's just a function of the -- you're kind of early in the auto cycle coming off a pretty steep bottom? Are you feeling better about the auto end market? I mean certainly feels like SAAR is still pretty lumpy, but just any quick thoughts on how you see the auto demand going forward.

I think there are 2 reasons. One, we went down more than the others. Our overall reduction post-COVID inventory was higher at the customer because Microchip was less flexible. So we have more inventory at the customers. Therefore, we went down more. So our recovery has been stronger.

And secondly, I think we're gaining share in pretty much all markets in the last 2 years.

Your next question comes from the line of Joe Moore from Morgan Stanley.

Great. On the topic of capital allocation, how are you guys thinking about M&A and inorganic growth at this point? Is that still something that you're it seemed like you were maybe through with that, but just how are you thinking about that now that there's a little bit more activity.

We are not thinking of any large inorganic growth. The only things on our plate are some small companies, technology where it could help a head in there like a halo is a very small acquisition that moves our road map substantially on AI on the edge products, which are products that we would have produced in the last 4 -- next 4 years in our business units and we get those day 1 but it's not a lot of dollars. And eventually, when we close the acquisition, you do know what we paid. But we're not engaged in any large acquisition today.

We're going to use the money to bring down debt. We are not comfortable with $5.5 billion in debt.

Makes sense. And then on the data center, I guess it'd be great to just at some point, get -- I appreciate you breaking out the numbers. I can get more of a sense of breadth. You talk a lot about the PCI Express opportunity, but I know the history of the company with PMC Sierra and Standard Microsystems and all of the IP that you have. I just -- not really a question, but just it would be great to get a sense of the breadth of opportunities that you guys face in that business at some point.

Noted.

Your next question comes from the line of Will Stein from Truist.

Great. Steve, I'd like to ask about the Defense end market specifically. As I recall, following 1 of your acquisitions years ago, there was a big supplier to the DoD. I think you're now the largest semi supplier to the DoD. And I think the reports about depletion of weapon systems has been pretty widely covered and there's an effort to rebuild those.

This has been a strong end market for you for a while, I think, and I think your outlook was strong there as well. But I wonder if, from your perspective, when you see the orders on the books, has the rebuild already been sort of placed in the supply chain? Or is that something that we should consider as possible upside more sort of on the come?

It's more on the come. The -- we are seeing only the front end of the orders on missiles and those kind of products. So we've gotten some orders to which we're delivering, that sector was up 45.6% in June quarter over the June last year. So some of it is in there. But the growth that the government and Department of Board is talking about on this thing is a production increase of 4x to even 4 to 8x is a number that we're hearing from prime.

But the primes don't have the capacity for everything else to ramp that much. So they're trying to get everything needed to really ramp that much. So a lot more growth is ahead. And I have spoken to the CEOs of the primes and they're asking me, are you ready to substantially ramp this and I said, ramp what? I mean you got to give me the order so I can figure out what components are needed, a general question, are you ready to ramp?

Yes, I'm always ready to ramp. So I think that's really where it is. They are trying to figure out their own needs regarding what all the things are needed. They're contacting the supplier, figuring out the ramp of capacity then give the orders and we start building. So our orders are strong, but it's only a fraction of really what is needed out there.

Appreciate that. If I can follow-up with a question about AI Data Center. I think what most companies have expressed through this earnings season is that, that end market still looks quite strong and stable I think you've said as such so far this call as well, although the stocks are sort of telling us that it seems like there are some people who are very concerned that the spend may roll over quickly?

Or maybe there's already well some fears being expressed, let's say. So what are you seeing in terms of the duration of the backlog, the stability of the backlog and outlook, are there perturbations that make you nervous about that backlog? Or has that changed for the better in that perhaps it's gotten even more stable and longer duration in the last couple of months, let's say.

Well, I would say if you take 2 pieces of it. One is our Data Center Solutions business unit and second, you take all the catalog products. On all the catalog products, I think we're seeing a growth even slightly higher than the Data Center Solutions group. Those are going from 288 to 500. And there are lots of new design wins that are still ramping and it's not -- it's single design, when it's hundreds and hundreds of them all over the place. They are in the rack, in the power, in coming into the data center, just all over, security products, memory products, parts on the card, parts from the power management, apart from the bus converters and all that kind of stuff.

So that is really just very broad-based and very difficult to get your arms around by design wins because it's just too numerous. When you get to the Data Center Solutions group, there, like we have talked about, 14 design wins, I think we have talked before that we were major suppliers on Gen 2, Gen 3 and Gen 4.

And we were very late to market on Gen 5 during the post-COVID period, and we essentially lost all of our Gen 5 business, we got none by the time we came to market that business was gone. Now since then, we produced our Gen 5 products, and we're getting lots of revenue in the second source position and it is ramping and Gen 6 is just starting a part went to production at the end of June.

So we're just shipping first parts to customers as we speak. So there, we're in the early inning and saw growth ahead because is coming from low numbers. Does that answer your question?

The indications that you're getting from customers, is the pattern of backlog or the pattern of orders, the backlog, the duration of it? Is it stable and getting longer? Or are there things that are moving around fast enough that suggests there's, I don't know, some lack of certainty, let's say?

And I'm saying that the customers, who are shipping those parts in volume today, they can probably answer that question better. We're not shipping in volume we have won 14 design wins on which we have shipped nothing yet.

I want to add 1 thing, and this is more in response to Joe Moore's question, a question ago. Sajid worked with our business unit leaders and produced a data center slide at Slide 14 on the investor deck that we posted on the website. I think you guys will find it interesting -- the left-hand side of that slide speaks about our Data Center Solutions business, which we've talked about more on a public basis over the last few quarters, but the right side of that slide talks about the other products and some of the areas where that plays.

I think that might help you, Joe, and others as you look at that slide, and you can obviously follow-up with Sajid, if you have any questions.

On the right-hand side of that slide, it talks about 5 specific areas. One is the grid to rack and in rack power. So that's really bringing power to the Rack. Second is a system management various Ethernet switches, USB hub controllers, MCUs that are doing predictive maintenance, fan controllers, temperature sensors, those kind of parts -- the third is timing and synchronization.

There is a tremendous amount of accurate nanosecond equitacy of timing needed and all the systems so they can all work together and providing -- we provide that timing and synchronization forage security, so you can't get into it and hack it and all that and 50 is networking, router switches, Internet price, optical modules and other stuff. So take a look at that slide, it's useful.

Our last question comes from the line of Joshua Buchalter from TD Cowen.

Let me echo the appreciation for the incremental disclosures. Maybe a follow-up on the last 1 as you brought up that slide. Any way you can maybe rank order the contributions today or even a year or 2 from now on the different buckets on that Slide 14 for Data Center?

I don't know about a year from now, but not today. We'll take that question and advisement and take a look at what we can do.

Okay. Year to try. Anyway, I wanted to follow up on the pricing increases, and I'm sorry for picking up this. But I was wondering -- any help you can give us on the scope? Was it mainly with channel partners, but was it channel plus direct customers more across different end markets? Or was it just across the board?

It was largely across the board, but there was not an equal increase on every product, and there was an equal increase on every customer because every customer buys a different mix of various different products. So there was some correlation to whether we're producing internally or externally where the cost increases are and there was some matching of the cost increase to the ASP increase of the product. So it was a very complex equation because we ship 100,000 SKUs, but it was very, very broad-based.

Well executed and generally well accepted by our customers, I would say.

I mean 1 distribution comment to me directly was that Microchip executed the best among many of their other suppliers who have done the price increase, we were the model. We would stand at how it should be done.

This is the conclusion of our Q&A session. I will now turn the call over to Mr. Steve Sanghi, CEO and President, for closing comments.

Well, thank you very much. I think we delivered a great quarter, and we are set up to deliver another outstanding quarter with a huge increase in gross margin, EPS and everything else. And there are several conferences this quarter, and Microchip management will be at those conferences, and we look forward to seeing you at those conferences and continuing this dialogue. Thank you very much.

Ladies and gentlemen, this concludes today's conference call. Thank you for your participation. You may now disconnect.

Welcome to the Microchip fireside chat. My name is Mark Lipacis, I'm the senior semiconductor analyst from Evercore ISI. We're very honored to and excited to have Steve Sanghi, the Chairman and CEO of Microchip; and Eric Bjornholt, who is the Senior Vice President and CFO.

So I think we'll just jump right into it. One of the biggest questions that we get is how big is any one of our company's data center and AI exposure. And you guys had a press release this week, which kind of spelled this out. You talked about 2 different -- 2 segments here, and there's a Venn diagram, and the data center and compute end market, which is 18% of your revenues. And then data center solutions business unit which was $300 million in '25 expected to grow 65% to $500 million in 2026. So Steve, maybe if you could help us understand what is the difference between these 2 categories that you broke out for us?

Yes. Thank you, Mark. I think many of the investors we have talked to were a bit confused and they felt that the $303 million that we broke out was the total size of our data center exposure, and that is really not correct. So let me take a chance to clarify it.

So the end market that we break out every year is data center and compute end market. And that is 18% of our total business. So when you do 18% of our calendar year '25 sales, that comes out to be $787 million business, that is data center and compute.

Now out of that $787 million, our Data Center Solutions business unit, which makes products 100% for the data center only market, was the $303 million, which leaves the $484 million unaccounted for. And some of that is data center and some of that is compute. And that is harder to break out because that has our catalog microcontrollers, our catalog analog products, mixed signal products, memory, security, timing, temperature sensors, fan control units, all that kind of stuff that goes in PCs, it goes in data centers, it goes into power supplies and they could be for data center or PC. So a portion of that $484 million is also in the data center. And therefore, the total data center exposure is a $303 million plus a significant portion of the $484 million.

Got you. So the $303 million that is products that you make specifically for the data center. So is it fair to say those are ASSP kinds of products or...

They are. They include PCI Express. They include storage controllers, storage accelerators. They include the storage memory products, PCI Express memory controller, CXL memory controllers, so all those parts, exclusively made for the data center. And then you have a huge amount of exposure to the data centers from catalog products.

Got you. Okay. So now you talk about a 65% growth for the -- for that business, the Data Center Solutions business, the products that are specifically made for the data center business, you didn't talk about the growth of the broader category data center and compute end market. Is it fair to say that, that in aggregate, that 18% of revenues would be a higher growth business for you than what you would expect for your corporate average. Is that fair to say?

So we didn't try to estimate the $787 million because $787 million is a mixture of it. A portion of those are microcontrollers that are going to laptops and loading docks and others and their growth rate is nowhere close to the data center growth rate. So in that, there are some high-growth products for the data center, but there are also some low growth products for PCs and docks and printers and all that. And we didn't try to estimate the overall growth rate of that. But year-over-year, our end market mix hasn't changed that much.

Fair enough. Okay. And then now in the press release, I believe you also discussed price increases. We do channel checks every quarter. And this past quarter, we identified 16 separate analog companies that took their prices up. But you were highlighted as one of the few companies that did not. So the question is, why did you wait so long to do so?

So it's a very good question. So I think there are 2 kinds of products in the industry. There are commodity products where they're easily exchangeable from one supplier to the other like the memory products. And on those commodity products, often you can increase the price when the supply is tight, and then purchasing manager will beat you down when the supply is excess and the parts -- the costs go down and the ASP goes down. But then there are second kind of products, which are proprietary products, which is what we make, microcontrollers, analogs, FPGA, data center products and others.

Historically, in our industry, prices have not gone up for the proprietary products. There is an unwritten promise to the customer that if you design with my proprietary products, you can safely think that I can deliver this price for the length of that design. COVID was the first time in my memory post-COVID where the shortages were so acute that people took advantage and raise the prices on all the proprietary products also. There were also a high amount of input cost inflation at that time. So it became necessary to pass some of those price increases. But that was really the first time. Historically, prices have always gone down, maybe low single digit.

So this is now the second time after COVID where some of our competitors have raised prices just because they can. Now we raised prices 3x post-COVID and we managed to damage our relationship with the customers. Some of them were upset. Pricing was not the only thing. We also had a very unpopular PSP program which force the customer to take the products when they no longer wanted the product, and then they got left with a year, 2 years' worth of inventory, which they have been bleeding for the last 3 years. And finally, it's coming close to correction.

So we came to an era of significantly, negatively affected damage customer relationships. When I returned back to this job in November of 2024 for the past 18 months, we have visited thousands of customers around the world and repairing our relationship, which I consider was a very successful project. And relationships are good. We're winning business. We're gaining market share. So therefore, we have taken all the input costs that have gone up so far in the first half of the year. And they're mostly internal cost, people cost, gold, copper, some raw materials and others. And we have absorbed them without giving a price increase to the customer.

In fact, despite that, our gross margin has gone up 950 bps in the last year. So we were pretty pleased with that, and we are in the striking range of our long-term target of 65% when while we were repairing the relationships, we took the high road and while others were raising prices, our customers were praising us how we were doing.

So now why the switch? So during this first half, while we had some internal cost increases, we absorbed them, we did not get external cost increases from our foundries and assembly test partners. We were able to hold them at bay. And finally, we can't hold them at bay. Now we have price increases with effective dates and percentage wafer cost or assembly test cost is going up in the second half of the year. So therefore, we felt prudent that we can no longer do that, and we needed to take those costs and pass it to our customers.

So what we're doing is we're taking the costs we have absorbed in the first half plus the new cost increases that are coming in the second half. We have totaled them up and we'll margin it up and distribute it among our customers. So considering we have absorbed the first half cost ourselves it will be slightly margin accretive, but the goal is really to be margin neutral. But the fact that it's a very complex equation across thousands of customers and 100,000 SKUs. We do not know how to hit the bull's eye, so we're going to hit on the right of it and not allow it to go on the left of it. So therefore, it would be slightly margin accretive.

Got you. How -- I appreciate that you don't want to give the specific variance even though I know I'm going to guess you have the variance, but how are you able to absorb the cost? What did you do to absorb those costs? And your gross margins went up despite higher input costs?

So we had very large inventory reserve charges. We have very large underutilization charges. I believe 3 or 4 quarters ago, seem to remember the number, we had $122 million of inventory reserves plus underutilization charges in 1 quarter. I believe that was a June quarter last year.

And as a business started ramping and as we cut production in the fab, our inventory charges came down by a huge amount and these internal cost increases was a small fraction of it. So our gross margin was rising, while we absorb those small amount of costs.

Got you. And what is the -- what is the impact to the customer relation shift? Is it -- as you held off raising prices, and we -- to me, it was remarkable to see 16 companies that we identified, and I'm sure there's more that they raised up prices. Does that drive -- does that help you win some share back or...

It did. It did. We have thousands of customers coming back to us. We've gotten praise from our customers, how we treated them. And we told them we were not increasing the prices, and they were very happy with it. So one, it helped us improve the relationship, which we already were doing it through other means. When the competitors started raising prices, it accelerated our customer relationship improvement, and we got -- we're winning new designs, we're winning every push because customers think, others are raising prices and we're not.

So the question obviously comes now as we also raised the price, how would that go? Number one, our price increase would be small. Number two, it will be very justified. We are passing on our cost increases to customers will understand. Competitors have raised prices, some of them more than once and we held them off for the first quarter -- first half, and we're going to raise some in the second half. Most of these will not be attractive until probably in September. So mostly the effect will be in the fourth quarter. So I think customers will really appreciate and it will not damage our relationship.

So -- and that's interesting for me because our checks indicated that those 16 companies were effectively raising them instantaneously. So it sounds like you're still like a 6-month window where you're not racing relative to the competition out there.

So on proprietary products, you can essentially raise the price and put the gun to the customers head and forced it. It doesn't mean you have to do it. You can do it, but it doesn't mean you have to do it. And we have decided not to do it.

And that, in part, at least, is driven by your desire to rebuild the goodwill with your customers?

Yes.

So -- and if we look at the data center solutions business, you mentioned the 3 buckets like storage controllers, PCI Express, CXL and switches and retimers. How -- like rough numbers, like are these equally sized businesses roughly? Or is there like a skew on one group versus the other?

They are equal size, wide brackets, but they're roughly equal, 1/3 in the storage controller area, 1/3 in the memory controller area, where we have the solid state disk drive controller as well as hard disk drive controllers and some CXL and then the PCI Express, the retimer revenue is 0, we just haven't even put the first part in production.

Got you. And then can we -- can you drill down on the PCI Express Gen 6 product that you have? I believe that you mentioned your market share on the Gen 4 product was pretty good. Gen 5, you missed a product cycle and so quite a bit disappeared. And now you're coming back with Gen 6, and you seem to be optimistic there. So can you help us like quantify like what did you lose? And what could you come back? And why do you -- where does the confidence come from?

So this business was acquired through Microsemi. Microsemi had acquired PMC-Sierra and then when we acquired Microsemi in 2018, we got that business. And PMC-Sierra/Microsemi had done pretty good on Gen 2, Gen 3, Gen 4. Actually, Gen 4 happened on MYCLOCK after 2018, between 2018 and 2021. So in all those generations, we had a good share of the market. It was very good business, high profitability.

And then during the COVID and post-COVID years. It was era after Gen 5 and Microchip was about 2 years late to production on Gen 5. Basically, we tried to design -- I wasn't the CEO at that time. I kind of stepped down, so it was a different management team. We tried to design SerDes inside. And every 2 weeks, it looked like we were 2 weeks away, and it kept on for too long and nobody really pulled the trigger and saying, enough is enough. We need to get the product out to the market and let's go license SerDes from somebody else, which is what I did on Gen 6, and that's what we're doing for Gen 7. But when we were 2 years late to the market, we basically lost the market, people designed it with whoever had the part.

So then eventually, when we came out with Gen 5 2 years late, we won some designs in the last 1.5 years. But they're all in the second source position because the pole position had gone to somebody else.

So what I have generally said is we lost several hundred million dollars of business per year on that Gen 5 alone. We haven't quantified exactly, but it was several hundred million dollars. So it's a very, very large loss and in that way, we can never allow it to happen again. And when we came on Gen 6, now we have one of the best parts in the industry, the only 3-nanometer part in the industry on PCI Express Gen 6. We now have 8 design wins. One of them is a $100 million per year design win that we have talked about. Every customer that we had in Gen 2, Gen 3, Gen 4, and including all the customers that we lost in Gen 5, they're all evaluating our chip because it's a better chip in the market.

Now the question is how many of those we will get back when we hurt them on Gen 5 because many of them were waiting on Gen 5. We made commitments, and we didn't meet and then withdrew the part almost from the market because we didn't have it. So we got some amending to do. And so therefore, we are conservative in the amount of penetration we can make on Gen 6. We think we should do better than anybody is saying, but we're not quite willing to say that yet.

Well, if you have 1 design win at $100 million annually, it seems like you're climbing some of that back, climbing some of that back.

But it seems like in the data center market, winning a $500 million design is not too much. So the $100 million uphill...

That would be a lot for me.

I'm looking for my $1 billion design win.

Right. So -- and can you help us understand is to what extent is this a product that ships into data centers that are an enterprise-oriented versus hyperscalers. I think there's a view that it's not a hyperscaler product per se. So what can you...

It is a hyperscaler product. So every hyperscaler, every server manufacturer and lots of semiconductor manufacturers that build boards by bundling their CPU, GPUs together with PCI Express. They're all looking at the part, and they're all either customers or potential customers of PCI Express.

One point I wanted to add is, over the last couple of years, 2 or 3 years, while it was the era of training the large language models, the GPU to CPU ratio was like 10:1. Now as that is shifting towards more inference, the GPU to CPU ratio is dramatically changing in the favor of CPUs. Most people are saying it's at least 1:1. I've also read people saying it's 1 GPU to 2 CPUs.

So therefore, when the -- in the data center, the percentage of GPUs to CPU moves in the favor of CPUs, it's a bonanza for PCI Express because PCI Express connect the GPU to CPU, CPU to memory, CPU to any peripheral, GPU to any peripheral. When there are lots of GPUs, the GPU to GPU is through NVLink. But GPU to anything else or CPU to anything else is often through PCI Express. So therefore, any estimates that we have seen so far on the size of the PCI Express market, and I don't have any numbers to quote, would be found underestimated. Therefore, it's a big opportunity for us.

So I will say, our own checks indicate when we talk to you about a couple of dozen hyperscalers that CPU to GPU ratio goes to 1:1 or 2:1 and that's kind of where they -- the average is landing between 1:1 and 2:1. But we also have learned this week that with agentic that the agents are calling more database workloads, and that's causing more demand for your standard CPUs also. So it's not just an agentic -- agentic CPU idea, it's agents caused demand for the standard CPUs also, which is quite interesting.

And those CPUs will require PCI Express to access memory, storage, RAID cards, peripherals, anything else.

Got you. Okay. So I want to shift to the aerospace and defense market, which I believe is one of your fastest-growing businesses. But was there anything else in the data center business you think is important for investors to keep in mind?

Anything else going on?

Yes, we covered it. Okay. Got you. All right. So this was, I believe, it made up 18% of your revenues in fiscal '26. Can you talk about the segments here? What's driving that?

So there are 3 segments in aerospace and defense. There is aviation. There is military hardware, offensive and defensive and then there is space. And there are things happening on all 3. For a while, the aviation production was way down when the MAX [ 750 ] planes were not really being manufactured. And now Boeing has the largest backlog ever in the building planes like crazy and -- and we are in other planes, too, French planes and others. And parts are loaded with Microchip's products on these planes. So that part of the business is doing very well.

When you look at the number 2, which is a military hardware and software, first it is the largest among the 3. And we used up about half of the arsenal we had built in 20 years and 2 months war. And so they're trying to rebuild it and Trump has asked all the primes to ramp their production 4x and to ramp production 4x, many of them have to build the new buildings and new factories. So this is not a short-term kicker, but it's a very long-term sustainable growth from the military hardware buildup.

And those primes are coming to you...

Those primes are coming to you. I've taken direct phone calls from CEOs or primes and saying, what is your capability to ramp up? And I'm asking, give me the order. They just said, are you ready? We're ready to do work. I mean we make thousands of products. You have to give me a specific product. How many do you want? Is it missile? Is it plane? Is it -- so the thing is that we are in every single military offensive and defensive hardware. We're in every missile, every plane, every battle tank, every rifle, every radar installation, every interceptor, every drone. And so it's very, very broad, and there are hundreds and hundreds of products that go into that structure. So they really -- you need to work either by a forecast or by order on specific parts. And we're getting those orders, but majority of them, we haven't gotten yet because it's still being worked, and they're really getting their own act together because our product is not the only product needed and it starts with ammunition and metal and housing and all that casings.

You need a complete kit...

Yes, you need a complete kit, and they've got challenges all over the place in being able to ramp it. So that part of the business looks very good.

And third piece of that is space. America has new fascination with space that I haven't seen since the Apollo years. We want to go back to moon and land the man on the moon in '28 and the recent Lunar mission where we went around the Lunar orbit, it was loaded with Microchip product. There are multimillion dollars of product in that mission and we'll have in every mission, I think there are 2 or 3 missions before we land the man on the moon. And then, we want to go to Mars and we want to colonize Mars and I don't know why, but as long as I'm making money. I'm okay with it.

And then you have the fourth piece of that, which is the new space. And these are low Earth orbit, the LEO, where SpaceX and others are trying to build a constellation of these satellites. And that market is a little different because they're not trying to send them up for the last 50 years like NASA is trying to do. If they last 5 years and come back in and run up in atmosphere, they just launched another 1,000 of them. They just do it in volume. They do it cheaply, and it's a different business model.

And many times, they're using lower-grade parts. They're certainly not using deep-space parts. Sometimes they're using the radiation-tolerant parts. Other times, they're simply using the automotive-grade parts and using them in parallel with a high amount of redundancies. If one fails, the other one will work and things like that. So we are participating in that, too, and we don't know how large that market becomes.

So when I think of this business, I think because you have to build in the redundancy or the -- and this -- the concept of rad-hard packaging, like I think, these are higher gross margin of businesses. Is that fair?

The space part of it, which is radiation-tolerant as well as radiation-hardened are very high gross margins with highest -- incredibly high gross margin. Because it very hard to make and it takes a long time to make them and then you have to build the parts and then take a sample and burn them in and test them with the radiations for a period of time, and then they pass, then you can ship a lot. So it's like a 9 months to a year production cycle. The parts that are going into LEO, especially if they're not radiation-hardened, if they're automotive-grade parts, then they're just off the shelf, they can be bought from distribution. So that model is different.

Got you. And do you -- some companies talk about -- we'll give a TAM per satellite? Is that something that you have shared in the past?

No, I don't know what that is.

Okay. Got you. Okay. So I know that you are a student of the cycle. And so I want to get your perspective on -- how do we think about where we are right now. Do you think your customers downstream from you, not just 1 layer deep but multiple layer deeps, where do you think they are in the inventory restocking cycle? Are they below normal? Are they normal? Are they above? Are they trying to restock? Where are we in the cycle?

So, we had a lot of inventory post-COVID because of the PSP program, we didn't allow customers to push out or cancel the parts they no longer needed. So as a result, we had nearly up to a year to 2 years of inventory, a lot of customers and distribution inventory was more than twice of really what would be normal. And we have spent the last 2, 2.5 years now cleaning up that inventory. The distribution inventory is now normal, actually, I could say, it's slightly below normal, but it's not perfect. Some distributors are below normal, some others still have a little high inventory. So it's kind of out of mix also.

And the customers is a very long-tail customer, 110,000 customers, hard to know who has worked, but seeing that thousands of customers who were no longer buying the part because they were using inventory and now they're buying the part and customer count is increasing by the 1,000 is a very good indication that thousands of customers' inventory is coming down. So they're reengaging to buy the part.

So I think that part of the cycle where the inventory correction happens is largely done, could have another quarter or so left. And then the customers are now buying to their consumption. Now at the same time, as they're buying to their consumption, volume has gone up significantly. We're up 34% versus the same quarter a year ago. So in that situation, and with the AI growth, AI is crowding out some of the assembly test capacity at subcontractors, some of the processes also.

So as those shortages start to materialize and work in the system and lead times start to go up, I think that will start the real upcycle of restocking, which in our business hasn't begun yet. Maybe it's begun in AI and all that, but automotive customers are not rebuilding inventory, industrial customers are not rebuilding inventory, appliance customers are not rebuilding inventory. Would they build in the second half? Maybe, but they're not building it yet because our internal inventory still coming down, customer inventory had just corrected. Some of them haven't even corrected yet. So it's early part to use the word restocking. I don't think it has begun yet or it will be done later.

But now what is happening on the top of the inventory-driven cycle is the innovation-driven cycle and innovation-driven growth is layering on the top of that. And you're seeing the innovation-driven growth in data centers we talked about. You have seen that innovation-driven growth in aerospace and defense we talked about. We're seeing the innovation-driven growth in automotive, we haven't talked about automotive, and we're seeing it in industrial. Industrial, as the onshoring is happening and factories are moving from China to here, they're not duplicating the Chinese labor incentive -- intensive factories in U.S. They are automating them. They're moving arms, their robots, there's conveyors. So it's really -- factories are really being retooled for a lower number of people working in the factory in a high amount of automation, which means they're loaded with our products with sensors and controllers and PCI Express or USB and connectivity and all over.

So there's stuff happening in the industrial sector. The things happening in medical, advanced medical equipment, is loaded with semiconductors or open phase MRI systems where you don't have that claustrophobic feeling that they push you in for 40 minutes inside that tube. All these innovation-driven growth things that we're seeing in the industrial market.

And the automotive market, what's happening in the automotive market is there are -- the unit growth is low single digit. But the content growth is much higher. We've got 81 chips in 100 Genesis. We have 61 chips in Mercedes S-Class cars and on, and on, and on. So the number of parts are multiplying.

And what's happening in the cars is, today, there are 6 or 7 distinct protocols for connectivity. There is a car area network, which is called CAN bus. There's a LIN bus. There's an entertainment bus that's called MOST. There's USB connectivity in every car. There is Ethernet. There's RS-232, there's RS-485.

Now when the car was in a connected car, then these protocols can reside in various modules themselves because they're not talking to anybody. But if you want to be able to press a button in the cockpit, like I can do my -- in my Tesla and saying, what is my tire pressure? In 3 seconds later, it shows a tire pressure of all 4 tires on my screen. Then you're connecting to every part of the car. And when you're connecting to the car and you have multiple protocols, then you need bridges to convert signal from one protocol to the other, so you can centrally access it on whatever your screen is connected to. And that is expensive and then it's a complex design.

So what automotive manufacturers are trying to do is converge all those protocols to a single Ethernet-based 10BASE-T1S protocol that runs 10 megahertz per second. It's fast enough for the automobile. It's not your office Ethernet wire, which is an 8-strand wire and can carry 10 gigahertz. The wire we're talking about is single pair twisted pair, 2-strand wire, and it can carry 10 megabit per second for the length of the car that is good enough. And all those protocols are converging to this single standard, and we're leading the standard, winning virtually every design. These are '28 production year starts, some in '27, more than '28 and '29. So that's the innovation-driven growth happening in the automotive.

So I think that's how I will describe the cycle where there is an inventory cycle driven growth, which is a normal semiconductor. And then there is just incredible innovation in these 4 segments, which seems like it's below normal innovation activity. It has really accelerated. And with the help of AI tools, customers design cycles are shrinking, our design cycles to deliver these innovative products are shrinking, time to write software is shrinking. So it's very productive delivery of this innovation that should drive revenue and be a very profitable revenue, exciting time.

So we have a cyclical dynamic here from my standpoint, our own -- we debated this yesterday, right? We still -- looks to us like you are still shipping below your trend line. And if your customers are shipping to demand, but they haven't started restocking, I would say that's consistent with our analysis. You have to have -- you have to bring your inventories back to a normal safety stock to get there.

And then there's a secular dynamic. And so it seems to you like you have an innovation -- you have innovation cycles on 4 dimensions. If I think about the past, I would say, oh, you have an innovation cycle in handsets and then -- or before that in PCs, but it seems like you have 4. And I'm wondering if you think about the industry, which has grown about 5% or 6% over time, if you have AI and automotive innovation and industrial innovation, is there a chance do you think about this opportunity for that long-term growth trend line for your industry or for Microchip to go higher than that historical 5% to 6%?

I mean, absolutely, we completed a 5-year strategic long-range plan. in January, putting all this understanding into it. And I asked our 20 business units to give me a long-range plan that you can sign in blood, means it's conservative. And they all did that, and 2 quarters have gone by, we're well ahead of what they put together, which means they were conservative. But even the conservative numbers that came up for the next 5 years, its growth was pleasurable. It was good. And it was much higher...

I guess, you're not going to share that.

No, we're not going to share that, but it's much higher than the numbers you talked about. And if that is conservative, that's even more music to my years.

So we have a little bit of time left. I want to ask -- I don't want to let Eric off of the hook. So I'm going to ask him one and I got one last one for you. So Eric, assuming we are in the place in the cycle that it seems like we're on the same page on. What is your -- what are you going to do with the cash, the free cash flow that you generate?

So obviously, we're going to continue to pay the dividend at the levels that it's at, but we are still really focused on paying down debt and getting our leverage down. We got into trouble in last cycle when the EBITDA fell so far. And with that, we had to actually come to the market with an equity instrument last March doing this mandatory convertible preferred.

So we don't want to find ourselves in that situation again. So going to maintain the dividend where it's at, not do stock buyback, generate cash, pay down debt and get the leverage down. So this quarter, we will end with net leverage below 3x. It's not where we want to be, but much improved from where we were, and I think we should make rapid progress through the rest of the year.

Got you. Okay. Fair enough. And then, Steve, last question. You, as much as any CEO, talk to investors, your shareholders. What do you think is the biggest disconnect or misperception about Microchip versus how you understand the fundamentals of the company and the opportunity?

Well, I think somewhere close to 60% to 75% of the time gets taken by the data center talk, and they just seem to not be able to get off that topic in most meetings. Automotive doesn't come up, industrial doesn't come up, aerospace and defense, when they get off data center, then start talking about SpaceX and others. So investors are very focused in these couple of segments. And sometimes they'd miss the beauty of Microchip's broad customer base, broad end market exposure into definitely 4 of these markets that are doing really great. Appliance market is kind of so-so. So I think that's -- sometimes they're going to miss out and sometime in any remaining time gets spent on trying to push us on why the margin should go higher than 65%.

I've heard that on the earnings call.

Yes.

So all right. Well, fair enough. Well, listen, we're out of time. Steve and Eric, thank you for joining. Thanks for sharing the insights about Microchip and the industry. Really enjoyed the conversation.

Thank you, Mark.

Thanks, Mark.

Okay. Good afternoon, everyone. Welcome back to the session of BofA Global Tech Conference. I'm Vivek Arya from BofA Semiconductor and Semicap Equipment research team and really honored to have the team from Microchip join us this afternoon. Steve Sanghi, Chairman and CEO; and Eric Bjornholt, Vice President and the Chief Financial Officer.

We will start with some opening remarks from Steve, then I'll go through my questions, but please feel free to raise your hand if you'd like to bring anything up.

With that, really warm welcome to you, Steve and Eric, really happy that you could join us at the conference. And Steve, maybe let me turn it over to you.

Thank you, Vivek. And before I begin, I wish to remind you that during this presentation, we may be making some projections and other forward-looking statements. These are predictions, and the actual results may vary materially. So I refer you to our filings with the SEC regarding some important risk factors about the company.

So what I'm going to talk about is 2 things, which were both in the press release yesterday, you must have read them, but just going to expand on it a little bit more. The first is the data center. So we have been getting numerous requests from investors to give them more information regarding our data center business. So the challenge is that in our end market breakdown, we have been breaking out the data center and compute as one of the end market, which is about 18% of our business. On a calendar year basis where we provided the information based on our calendar year 2025 revenue, that 18% comes out to be about $787 million. There's plus or minus 2% error band because there's a lot of suppositions and estimates in there.

Now out of that business, the challenge is that we have a data center solutions business unit, which is pure data center, 100% of their products going to data center. And that business unit produced $302.7 million in calendar '25, out of that $787 million total.

Now that doesn't mean that our data center business was only $302.7 million, there are a lot of additional products, some of our microcontrollers, analog, timing products, security products, memory products and others, which go into multiple markets, data center being one of them, and many of these products are catalog products that also get sold through distribution and contract manufacturers which will bundle the total sale in a given product, some portion is going to data center, but some could be going to industrial. So there are a lot of estimates involved.

And our total data center business is somewhere between that $303 million to $787 million. But since there is much more difficult because of going into multiple markets. What we decided to do is give you our revenue for just the Data Center Solutions business unit, which is 100% into data centers. And that was $302.7 million and expected to grow to about $500 million in calendar year '26.

Now we'll continue to do more work on it and see if we can take the difference between $787 million and $303 million and break it out further into what is in compute and what is in data center, but we were not able to do that by yesterday. So that was the one part of the press release. And as I had various one-on-ones, I think some people were still confused where they thought the total size of exposure to data center was $302.7 million. That is not the case. That is only the data center solutions business unit.

The second item in the press release was pricing. So until last week, we were essentially saying we're not increasing prices to our customers. Essentially, some of the operational cost increases we have incurred with people, chemicals, gases, transportation, energy, gold, copper and others, so far, we have absorbed them without giving a price increase to our customers. And despite all that, our gross margin in the last year has gone from 52% to 61.5% in the last quarter, and we're guiding to about 63 point -- I don't know what the exact guidance is this quarter, but...

62.75%.

62.75% this quarter. And that's happening despite us absorbing some of the cost increases. And so far, we were able to push back on our suppliers, foundry and OSAT and other suppliers that we can take a price increase and we didn't incur them. But now it's becoming much more difficult because essentially they have been able to pass those price increases to all of our competitors, and they're passing those price increases on us and there are dates and times and percentages by technology, by foundry, by OSAT, that will be coming up where our costs will be going up. So finally, and we've been doing this analysis for some time. But a decision was finally made 2 days ago, Sunday evening in a worldwide call where we put all this data together that we have been putting it together for some time and made a decision that we will have to increase prices to take the cost increase we're incurring and margin it up and give it to our customers.

Now at this point in time, we have no estimates on the amount of cost increase or price increase on the customer, which products wIll see it and which customers will see it, how we will deal with distribution, how we will deal with direct customers. All that work has to be put together in the next 2 to 3 weeks. And then our salespeople and business unit leaders will take those price increases to our customers. So we don't really have any more detail on that.

We will not be telling you the average price increase for obvious reasons, one, that's a competitive information. And second, if a customer knows what the average is, then nobody will take more than the average, and that would create a problem if we want to exempt some customers where we are still repairing our relationship. So we had to make that up by giving more cost increase to somebody who's a non-partner. But if you come up with an average number, then nobody will allow you to do that. So we will not be disclosing how much price we are increasing and just executed in the marketplace.

So with those 2 updates, I'm happy to take any questions, Vivek.

Thank you, Steve. So maybe continuing on that pricing theme. As you mentioned, Microchip has -- by being somewhat more resistant to taking pricing, right, whereas many, I would say, in fact, all of your, in analog and kind of power semi competitors have taken pricing. How much of that, Steve, was just because you were still in the customer relationship repairing mode. And now when they see this, what do you think is going to be their reaction, does it impact that progress that you've been making?

So I think you have to look at through cycles. So in the last cycle, we took pricing up a little too much, one time longer than probably they did, and we were much harsher on our PSP program in the noncancelable, nonreturnable, nonreschedulable, then our competitors were. So we upset some customers where our competitors treated them better.

Now fast forward in the last 1.5 years, we have done very good with our customers, repaired it lot of relationships. And this time, our competitors took the very first opportunity and raised the price on them, while we told our customers that we're absorbing the cost increases and you're not getting a price increase. So net flow of the customer has been towards us. We've won many, many designs and customers don't like our competitors today, they like us. And some of them have done 2 price increases already. We haven't even done one.

So when we go back to them saying we have held a price increase now all year, and we've treated you better and all that, and they have seen 2 price increases from the competitors. So when they hear from us that we're doing a price increase, number one, they will not be surprised. And if the amount is reasonable, I think they would be understanding our relationship will still be very good.

Got it. And just to confirm, this is not happening because of shortages, right? I mean, your lead times are still, right, reasonable.

It has nothing to do with shortages or excess. It has nothing to do with increasing the price because we can. I mean if we really wanted to, we could increase the price more. So it's just purely have to driven where we don't want to gouge the customers but we don't want a headwind on our gross margin. So we'll take an estimate of our cost increases, margin it up and give it to them and be able to explain it.

Got it. And historically, Steve, has this resulted in some kind of prebuying or extra buying, anticipating that, oh, look, there is this another price increase coming? So let me buy more now? Or do you think it could be...

So it depends on how you implement it. If you say we're going to -- if you wait too long, give them too much of a time to buy the product, then they will try to buy some product in advance at the lower prices, we will increase the price on the backlog. And any incremental orders we see over and above the run rate, we'll schedule them at a new price after the price increase was effective. So therefore, there will be no incentive. And if anybody places the order, we'll schedule them after and then we are at the new prices. So there should be no games really. I mean it shouldn't affect our revenue. There should be no pull-in effect because we get to schedule it, and we'll -- just based on the lead times, today have 7 to 8 weeks, the window will probably give them will be shorter than that. So essentially, all that will fall outside of the price increase.

Got it. Makes sense. And then is there a scenario where this is a tailwind, albeit a modest one to gross margins? Or do you expect it to be neutral? Or how should we think about the benefit?

So a couple of things. We have absorbed some of the internal operational cost increases thus far. And when we're increasing the price, we're including in it. So there is a little bit of the recovery that we have incurred the cost and I will pass it on, which will make it slightly accretive.

The second part is, our target is to be margin neutral, but you can never hit the bull's eye and I don't want to have the risk of being on the left side, so we want to be on the right side. So given at all, it will end up being slightly accretive, but only because of we can't be so precise.

Got it. This kind of bridges into the next few questions about your data center. Is it fair to think that most of your data center sales are coming from parts that you get from foundry. So if that is the fastest-growing part of the business that you are always kind of exposed to this inflationary impact?

Well, inflationary impact is not only on foundries. We are seeing the similar inflationary impact inside because foundries, their cost is going up on their people and chemicals and gases and energy and transportation and all that. And that shows up in their wafer costs and they give it to us a higher price. But our internal manufacturing is doing the same. They have the same inflationary cost increases.

But in answer to your question, our data center solutions business unit products are all outside, but not our total data center exposure, some of our analog products, our microcontroller products, some of our security products, timing products that go into data centers are produced inside.

Got it. Okay. And the growth rate that you described for the data center business, I think, 60% plus. How do you benchmark that, right? Do you think it's the right growth rate because there's such a wide range of growth rates that we see across the data center, right, is like can that be sustained over the next few years? How should investors think about just your data center? First, maybe let's start with what is in that data center number? What specific end markets applications and then growth rate?

So in the Data Center Solutions business unit, we make 3 product lines with the fourth one coming. And the first product line is our storage controllers, storage accelerators and storage expanders. And in that product line, we have about 3x performance advantage to our nearest competitor. So it's very state-of-the-art and it's doing very well.

The second product line in that is PCI memory controllers and CXL memory controllers. And there again, we're kind of leading the industry.

And the third product line is PCI Express Switch. Now in PCI Express Switch, we did very well when we bought the business from Microsemi, they had done the Gen 2, Gen 3 and Gen 4. I'm sorry -- Gen 2 and Gen 3. Gen 4 happened after we had bought the business and Gen 4 happened on MYCLOCK when I was CEO last time, and we did very well in those.

And then we screw up on Gen 5. That didn't happen on MYCLOCK but it happened under Microchip Clock. And we were 2 years late to the market on Gen 5. And therefore, all those Gen 5 designs went to the competition. When we eventually came up with the Gen 5 product that again happened in the last year, we won a lot of the designs again, but we got a fairly small share in the second source position, not in the primary position.

So the commitment we made is that we've got to regain our Mojo, and that's why we produce the Gen 6 product on time on one of the best products in the industry, the only 3-nanometer product, significantly higher performance than the competition and 30% to 40% lower power.

So that part is winning designs like crazy. Every single customer who did business with us is reengaged. We have publicly said in the last conference call, we have won 6 designs. And now we have won 2 more. We have 8 designs on Gen 6 Switch and we have won design on the retimer. Retimer doesn't go to production until January. So this is a very, very early win.

So I think if we had not lost all that on Gen 5, then it will be a little bit easier to make assessments of how fast it will grow or what will accelerate. We know the market is very large. And the upside is just totally incredible. If you keep winning these designs, they go to production next year, we can acquire the capacity then this could be very big. But I think we are cognizant of the fact that we are coming from behind, having missed a Gen 5, then. And we're competing with a formidable competitor with a B name in it. And so we are just a little cautious in trying to forecast anything.

Now, one of the things that's happening on the Switch side is, and you may have heard about it or read about it, during the training phase of large language models, it was the era of GPU. And the GPU to CPU ratio was as much as 10:1. But now as you convert from training phase to an inference phase, it is favoring the CPU and GPU to CPU ratio is going as much as 1:1, and I'm even starting to read 1 GPU and 2 CPU. And PCI Express connects the GPU to CPU, to CPU to CPU, GPU to memory, CPU to memory, CPU to anything, RAID card, memory controller, GPU. So as the new industries calling it CPU renaissance, if there is a new CPU renaissance and you're seeing Intel's capacity constrained, others are too. And as the CPU volumes explode, it is very, very positive for PCI Express. Because GPU to GPU connection is through NVLink. But GPU to CPU connection is through PCI Express, which is standard industry protocol.

So the market strength and conversion from training to inference favors PCI Express. So if that market grows more than anybody expects and we should really get a fair share of the market. And I think demand is going to be strong enough with probably overall TSMC capacity limited that everybody will get their fair share to what the capacity they can acquire.

All right. Is there some simple math, Steve, that you have in mind? So if, let's say, the CPU market, right, people used to talk about $60 billion, then they went to $100 billion. Now the new number is $200 billion. What does that tell you about kind of the PCI opportunity associated with it?

It is probably calculated, I haven't done that yet. I think that CPU renaissance is relatively new. So I don't think I have seen the PCI Express estimate because of that change. And that will only describe the TAM, total available market. It still wouldn't -- still it wouldn't tell us what we could win. Like I said, we are making up the loss.

Is it as simple as your share of 3-nanometer allocation at TSMC? Is that kind of the first order way of looking at your share or that other...

What we need on 3-nanometer is 0.1% or 0.2% of our [ Jensen ] need. So I'll just tell them and give me all I need and [ Jensen ] -- may feel the change and it's working so far.

So is it deterministic what your share is? Or do you think that...

It's not deterministic. We -- the part goes to production at the end of this month and winning 8 designs before the part is in production is kind of unprecedented. So because the part is that good and it's doing that good in customer tests when they put them in the board. So we think we have a winner on our hands, but we are shy in forecasting until we actually start to see these wins going in production and then multiplying like crazy.

So is there a scenario where these kind of growth rates are sustainable? Or one could...

I hope they're not only sustainable, but they accelerate. But I'm not really willing to put a number on that yet.

Okay. Understood. Next thing, Steve, is just kind of the broader analog cycle, right? Where are we -- usually, these cycles have sort of the inventory replenishment phase and then the end demand improvement phase, right? Where are we? Just because the size of these markets, especially in the data center are getting so big, so it's hard to separate out a cyclical recovery from what are kind of the secular drivers.

That's a very good question, and I've been talking about it. So the first phase of our growth in the last year has predominantly come from customers and distributors that were buying product well below their consumption level because their inventories were high. And asymptotically, they're trying to reach where they're buying equal to the consumption level. So that fed the first part of the growth, which is typically inventory depletion driven semiconductor cycle growth.

But what's being layered on it now is, 2 other things. One is the new design that we have won in the last 2 years and even in the data center, since we were so late to Gen 5 PCI Switch, we lost all those designs. But when we came back with it, we won a lot of designs where we are in the second source position. And so that PCI Express Gen 5 is ramping, waiting for the Gen 6 production to start. So we're getting some growth because of that.

But now the third phase really is the innovation-driven growth. And there is innovation happening in 4 of our end markets. It's happening in, obviously, data center. It's happening in automotive. It's happening in aerospace and defense, and it's happening in industrial. I'll spend a minute on each of them.

Data center, we have talked about, lots of innovation going on, it should grow. In the automotive, so number of automotive units per year grow in low single digits. So that has -- never been the story. The story is the consumption level inside. On some of the high-end automotive cars, we have 81 chips from Microchip in one car and 63 chips another car. So there's a large amount of Microchip content in the cars.

And today, there are so many different protocols inside the car. There is a car area network. So there's a CAN bus, there's a LIN bus , there's an entertainment bus called the MOST bus. There's USB connection inside, there's Ethernet, there's RS232.

And in the past, all these things don't need to talk to each other. Your microcontroller or tire pressure sensor. It doesn't need to talk to anything. But today, you could press the button and say, what's my tire pressure? And it will show you picture of your car, full tires, entire pressure in each of them. My Tesla does that.

So then now the car is talking to every node that is outside in a car. So therefore, in the current car setup, you have to go through bridges to convert from USB to Ethernet, car area network to USB or through Ethernet, MOST bus to Ethernet. And these bridges are expensive and bulky and extra chips that need to be added. So where the cars of '28 and '29 and '30 are going is to consolidate all those protocols into a single Ethernet-based network requiring no conversions, everything is basically that.

And we are leading that charge. That standard is called T1S, which ran on a single-twisted pair Ethernet wire, not the 8-strand wire you have in your office, but 2-strand wire, single-twisted there. And that should drive significant growth and -- there's a lot of innovation happening on it.

So we talked about data center, automotive and quickly, the third one is aerospace and defense. So aerospace and defense has aviation, has military hardware and space. And they're growing in all 3, like Boeing is building planes like crazy again where MAX 7 production is back up. Military hardware, they're trying to replenish all the hardware they have used in the wars and Trump is asking primes to quadruple their production, and that's a long-term growth for us because primes are adding capacity with brick and mortar and all that, and that revenue should do very well.

In space, U.S. as well as the world has a new fascination with space. The fascination that hasn't been seen since the Apollo days. And so we should do very well with all our space content.

And the fourth final market is industrial with what's happening in industrial with onshoring, when the factories are being onshore. They're more highly automotive. They're more highly automated. There's a lot more robotics in them than the factories. They're moving from in China. So there's more content from us. Sensors, microcontrollers, PCI Express and all these parts in the factory. And then all our parts going into humanoid and robotics and moving factory line and all that. So I think all that is innovation-driven growth, that's leading on the top of the inventory driven growth, and therefore, future looks bright.

Got it. One maybe a question on margins. I'll bring Eric into the discussion. So gross margin, right, you mentioned 62.75%, right, in June. The target is to get towards 65%. I imagine most of that is just an underutilization, right, kind of recovery?

Let me have Eric address that.

Sure. So last quarter or the March quarter, we had $46.6 million of underutilization charges. And if you add that back to our gross margin last quarter, we're essentially at 65%. So it's just going back into our capacity. It's not all going to happen at once. There's going to be a steady state that we're increasing production in the factories. We're essentially ramping our Oregon factory as hard as we can this quarter. We are shipping at a level that is well above what we're producing at. And so that's why inventory has been coming down, and we want to make sure that we bridge that gap so inventory doesn't get too low. So those underutilization charges will get much, much lower by the end of the fiscal year.

Got it. So then what about the OpEx leverage side of it? Is there a certain revenue run rate where you can achieve your target operating margins also?

I think it very much depends on the slope of the revenue curve. Today, we brought back a lot of programs that went away when we kind of went through the doldrums of the downturn. So bonuses are back to high levels of our variable comp programs, merit increases are back for employees. So there's going to be growth in OpEx dollars, but a significant portion of that happened over the last 2 quarters as we are bringing these programs back. So this quarter, we are guiding to OpEx as a percentage of revenue at about 29%. Our long-term target is 25%. And you should expect as revenue grows, that OpEx is going to grow at a slower rate as we drive towards that model.

Got it. Last question, Steve, maybe on -- we always kind of look back at history and say, look at the last peak of Microchip was here, right? Now you're still like 30% below that. Many of your peers have caught up. Now your last peak was kind of unique because of its own characteristics. What is the right way to kind of frame what your growth trajectory? Like is looking at the last peak even useful, right? What is the right way to think about that?

It is not useful. In the normal past semiconductor cycles, you could do the analysis of area end of the curve, you ship to higher and then that has to be below and when you consume it, then it returns back. That analysis is much more difficult to do from the last cycle because when we were shipping at the peak of the cycle, we were shipping hundreds of millions of dollars into inventory.

And so last peak, really, I mean, it was GAAP revenue by the definition of GAAP but it was not going into consumption. It was going into inventory. Customers, in many cases, didn't want the product, because of the LTSAs and PSP and all that, we jam that product down their throats. So therefore, that peak was very artificial.

And secondly, in the history of semiconductors, other than the memory, there haven't been price increases, right? Microcontrollers, analog, you don't go through price increase cycles -- price goes down low single digits every year for the last 40 years. So therefore, the analysis is apple-to-apple.

When there are so many price increases in the post-COVID cycle, then when you do the revenue-based analysis, it gets distorted, you've kind of got to -- go to the unit-based analysis and even in a unit-based analysis, the mix shifted a lot because during post-COVID, people gave on the low end of the businesses. So they can take the limited capacity and give it to the high end of the businesses, in high end of the businesses, there are more dollars but less units. So the unit-based analysis becomes very difficult. So I think pretty much in the last 1.5 years, every analyst or investors that have done that analysis on Microchip was really disappointed because that analysis didn't work. And we told them that it won't work, but nobody listened. So don't do that.

What is the right analysis, then?

There is none.

So how do you know whether you're at the right timeline?

How does it matter? We're growing from here and take the number and look at the opportunity ahead of us with all these 4 markets I talked about, and that's where it is. I'm already saying that the inventory depletion driven growth. It pretty much comes to an end in the next quarter or so. So rest is an innovation-driven growth. So what was the past peak? I mean, how does it help you?

Right. Makes sense.

It only confuses you.

Right. Okay. On that note, thank you so much, Steve. Thank you, Eric. Really enjoyed that discussion.

All right. Good morning. Welcome to our -- the second day of our 54th Annual TMT Conference. Really excited to be joined by Steve Sanghi and Eric Bjornholt from Microchip. Gentlemen, thank you so much for being here with us today.

Thank you.

Steve, we had a really interesting discussion at dinner last night, I think, about why -- how unique this past cycle was. And it feels like we're coming out of it and we can finally maybe stop talking about it.

But as we reflect on, again, coming out of the cycle and things sort of normalizing, what were things that you think is different about this cycle and perhaps are going to change for the industry going forward as we think about future balance between secular and cyclical growth?

So I think the post-COVID cycle was an extremely unique cycle. Prior cycles were driven by like we had a PC cycle years ago, we had a cell phone cycle. We had an Internet cycle. So there have been various different cycles and semiconductor inventory cycles. But the last cycle just was so extreme where artificial demand became so high and it turned out at the end that a lot of the demand that the customers asked for were not real.

And when the customers' business came down, it created a huge vacuum almost in the air pocket for almost every company. And it was almost the worst for Microchip because we had a program, which was very inflexible, which took customers' orders for a year and wouldn't take cancellations or push out and that prolonged the downturn for Microchip.

So that was the unique in that cycle. Now where we are today is for the past 2 years, the shipments to customers have been lower than what they have been using and our shipments to distributors have been lower than what they have been selling out.

And that gap has been closing. And we believe that last quarter, the shipments were distributors and their sales out were pretty close. It was only a few million dollar gap. A long tail of hundreds of thousands of customers is much harder to tell.

But by just customer count, our customer count is increasing by thousands of customers, meaning that they had inventory, they were not buying products and they're starting to buy products again.

We don't know every customer's inventory, but the customer count is increasing, indicating that a lot of the customers are coming back. So I think in another quarter or so, that part of the cycle normalizes. Customers and distributors are buying what they're using and that part of the business is normal.

But then what you have is the innovation-driven cycle that is creating the growth. During post-COVID, every design engineer was working on substituting parts, whatever they could find. And for a couple of years, there was not so much new design activity. When the parts became readily available in a down cycle, engineers went back to work to design new products and went back to innovation.

And you are seeing now the front end of all those designs going to production in industrial, in automation, in aerospace and defense, and automotive and data center. So we're seeing a huge amount of innovation-driven growth. And I would say, in 4 main markets, end markets.

Clearly, data center, we're seeing it in industrial. We're seeing in automotive. We're seeing it in aerospace and defense.

Okay. I want to get to those markets, but first, I want to basically get through the cyclical discussion. As we think about inventory levels, like we like to think about the 3 levels of on books, channel and downstream. It sounds like you're very comfortable with inventory on the balance sheet, at least in dollars, not necessarily days and channel is lean and end customer inventory levels, you're cautiously optimistic that's 90%, 95% done.

Is that the right way to think about it? And that in a quarter or 2, we will basically be shipping to consumption?

Yes.

Okay. Got it. And Eric, is that correct that like you guys -- again, the days are a little high right now, but you want to get back into that 130- to 150-day target on the balance sheet just by basically revenue growing into the inventory in dollars?

Yes. So our inventory days peaked at like 266 days a year ago, March, and we've got that down to 185 days as of the end of this last March, and that includes about 15 days of what we call last-time buy inventory from foundries that have end-of-life product that we bought product for, for the next 10 years.

So if you exclude that 15 days, that's on 170, we're growing at the midpoint of guidance this quarter, 11%. Days are dropping very rapidly to get in that 130 to 150-day range that we've set.

Okay. Last topic on the cycle, I think. So a lot of your peers have talked about raising prices more recently in light of rising input costs, but I think also just because the demand has been so strong. I think given what you mentioned regarding the PSP program, you guys seem a little bit more reluctant to raise pricing opportunistically, certainly.

Can you speak to how you're strategically thinking about pricing right now?

So there are 2 kinds of price increases. One is increasing the price opportunistically because you can when the supply is tight. Historically, in the industry, that has happened in memory and other commodity products. Design-in products, microcontrollers, FPGAs, processors and others have really in the industry not followed gouging the customer on a proprietary product.

So that is a price increase we are less interested in because our margins are good, margins are still rising. We're going to get to a model fairly soon as the underutilization goes away. So the opportunistic price increase to gouge the customer to have your margin, I think that's probably not a good thing for us to do. But the other piece is if our costs go up by our foundries and our assembly and test suppliers and we're paying higher wafer costs and this and that, then passing those to the customer is something we're more comfortable with.

Now in the history, semiconductor prices have never gone up on that. COVID -- post-COVID period was the first time where inflation just went so high and everybody landed up raising prices multiple times in those years. And this is the second time it's happening.

But I mean, largely, we have been sending the price increases and succeeding so far from our partners. But if we are not able to do that, then we'll be willing to raise the prices.

And are you guys thinking about this as an opportunity to lean in and gain share as your competitors raise prices? And how are...

Because of the first factor, where I said, if the others are raising prices opportunistically, we're not. We're having lots of their customers come engage with us.

And I guess on that note, I mean, one of your key priorities with your 9-point plan when you came back was repairing customer relationships that had been impaired partly, I think, because of the PSP program. Where are we, I guess, in that process? And is this an opportunity to, I guess, accelerate that?

So that process is largely complete. We had a massive program where we visited thousands of customers around the world. And that part of the process has gone to normal. If you were to do an unexpected price increase to gouge the customers, we'll be hurting and reversing that process.

And while our competitors are doing that, we're taking advantage of it. But the repairing part is kind of complete. I mean you have to always work with your customers and visit them and talk to them. But there's not a massive repair program that we put behind us.

Okay. Let's shift to some of the secular pieces. I think the one that's certainly on investors' mind the most for the industry, but also increasingly you guys is the data center piece. Can you walk through -- you've got products specifically for data center, but then also your broader product portfolio that is applicable to data center.

Can you walk through the key exposure and what secular opportunities you're most excited about in data center?

So we have 2 kinds of products going into data center. The first kind of the ones which are the products we design for data center and sell only in data center. That number is readily available and it's clean. And that comes from our data center business units that build 3 different product lines.

They build storage controllers, they build the memory controllers and they build PCI Express. And the business in them is about 1/3, 1/3, 1/3. And the fourth is a retimer, but it goes together with the PCI Express, so I put them in the same product category from a product line standpoint.

All 3 parts of those businesses are growing significantly from calendar '25 to calendar '26. And those are pure data center products. They only go in data centers, and that part of the business is doing very well. What complicates the data center exposure for us is all of our other catalog products coming from microcontroller business units, analog business units, security business units, power management business units that also go into data center, but they also go into the other market.

So take a digital power supply product that we could ship it to a customer that builds power supplies. They build the power supplies for PCs. They build it for servers. They build it for telecom equipment racks. And they don't break it out for us how much goes away, which is what makes this breakdown so complicated.

Similarly, we make security chips that secure access to PCs, access to cell phones, access to other communication gear and access to servers, which again is multi-market product. I can take a very simple product like a temperature sensor. The temperature sensor can go into a thermostat, which is a consumer product.

It can go into a PC to control the fan. It can go to a server also to control the fan or you can go to control an air conditioner. Now you got data center, PC, industrial, home. This is what makes the breakdown very, very complicated.

We have a substantial exposure to data center, but not yet comfortable in breaking the numbers in a way. We could make some wild guesses, and we could -- we're trying to get some more data from our customers to help us break it out, but our customer base is also very broad and long tail.

So any analysis would have some level of plus or minus around it. And we're working on it, and we want to share more information, but we don't have it today.

Maybe we could double-click on some of the more data center-specific products. And you're making a bit of a comeback with your PCIe Gen 6 switch. Can you talk about your legacy and heritage in the PCIe switching market, how you got that IP, why you lost some share in Gen 5 and then why you're optimistic that you're going to regain share in Gen 6?

So our data center business unit was acquired as part of the Microsemi acquisition in 2018. And we did -- they had done Gen 2, Gen 3 before and Gen 4 happened on our clock after 2018. And then Gen 5 happened during the post-COVID years.

And Microchip misexecuted on the Gen 5. Our part was nearly 2 years late to the market. And so therefore, all the designs went to the competitors. When we eventually came with the Gen 5 part, we got some of the designs, but we only got a second source position.

So we're shipping today Gen 5 in volume to a lot of customers, but much smaller volume than we were shipping Gen 4 and Gen 3 because we didn't have a primary position. We had a secondary position.

So we made a very strong commitment to Gen 6 and Gen 7, and we leapfrogged on Gen 6, where our competitors' parts are on 5-nanometer technology. Ours are on 3-nanometer. We have 160-lane device. Competitors have 120-lane device.

We have 30%, 40% lower power, significantly higher speed and more lane count. So we have a leadership part that is currently being looked at every hyperscaler, every enterprise, every server manufacturers, and we have disclosed 6 design wins, not by name, but there are many, many others who are looking at it.

So our knowledge of PCI Express and our knowledge of these customers and their knowledge about us is not new. We just executed one generation. And as we have a leadership part, it's very well recognized.

It's been written about. The companies who have design win with us, they have talked about it. They've given us testimonials, which we have shared with other customers. So I think our recovery and comeback on Gen 6 after the Gen 5 file would be very, very strong.

And I guess, can you talk about the competitive backdrop? I think the market has certainly gotten a lot bigger as data center infrastructure needs have grown. I guess, one, are you primarily competing with general purpose servers or accelerated server racks? And how has the competitive environment changed since you were more dominant in Gen 3 and 4 versus now?

So PCI Express is used in service and data access, GPU to CPU connection or CPU to other things connection. And as the data center market is exploding, the PCI Express demand with that is exploding. It can be in a scale up and scale out. So you're building new servers, you will need PCI Express. If you're replacing racks and putting high-performance things, which is kind of scale up, then you will need the high-performance PCI Express in that also. So the limitation of memory to CPU, memory to GPU doesn't become a constraint. So we're really participating in both scale up and scale out.

Okay. And then you talked about the 6 design wins. I think you've said that one of them alone is over $100 million. What's the time line in which we can expect that to sort of meaningfully start to layer into the model?

The device goes into production at the end of this quarter, which is another month, and then we start making a small amount of shipments, but the major ramp is next year. And the $100 million design we talked about will be production next year.

And you also recently announced your entry into the PCIe retimer market. Can you talk about where that IP came from? Why the concerted effort to move into that market? And then how synergistic is it and important with your PCIe switching business as well?

So we have some lower performance retimers in other markets, communication, et cetera. We didn't have a data server class retimer that will go with PCI Express in a very high performance, rebuilding the signal when it deteriorates. So this is our first retimer product for the data center market. And it works in conjunction with Gen 6 because as the signal weakens, retimer rebuilds a signal and provides a strong signal to go forward.

We basically did it ourselves. It's the same resources, data center business units, the design and application engineers that build the part. The part is -- has been evaluated by customers. We have one design win on it. Part only came out of fab about 6 weeks ago. And it works beautiful. I've seen it, and it works great, and I think we'll be very successful co-selling it along with our PCI Express...

Okay. One more on data center, maybe it's actually not data center. Earlier this week, you announced a 3.3 kilovolt silicon carbide module. I hadn't historically thought of Microchip is playing into that high-voltage power discrete and power semis market. Can you provide some background of what this part is targeting? And maybe your overall exposure to high-voltage power?

Yes. So we started a silicon carbide business quite a while ago, and we had grand aspirations in the silicon carbide when every EV was going to use it, there was a mandate nationally that by 2030, all the cars will be EV. The total size of the silicon carbide market was expected to be just very, very large.

And that's when we started the effort in the silicon carbide market. We made some FETs. We made some lower voltage 700 volt, 1,700-volt devices. This is the highest one. Then 2 things happened in the last 5 years. One, I believe it was Tesla that first came out and said that 80% of what we want to do in electric vehicle with silicon carbide, we can do it with silicon.

It's not as good as silicon carbide, but it is good. And that brought the market down significantly and crashed the whole silicon carbide. Second thing that happened is the EV mandates went away and consumer largely did not adopt EVs. They more adopted hybrids and range extenders and other things and the rate at which the EV market is growing significantly slowed down.

So there was a second blow to the silicon carbide market. And the third was the Chinese overbuilt for capacity in silicon carbide. So therefore, when I came back as CEO 1.5 years ago, in my deep dive analysis, we largely took the emphasis on silicon carbide away because of those reasons and combined it with our silicon power module business, so we're doing some silicon carbide, but the business unit resources are doing a lot of other things.

So this is one of the device that came from them. It will go into EV charging infrastructure. It could go into a data center. It could go into a power management where you want to do really high-power switching without loss of power, without heat generation.

And is this -- how should we think about, I guess, the time line for this materiality compared to the PCIe business, I guess?

This is not going to be a huge business.

Okay. Got it. Shifting gears, Aerospace and defense is, I think, an area that investors appreciate your position and high exposure, both from legacy Microchip, but also a lot of the Microsemi assets you acquired. Can you walk through some of the key products and exposures you have within aerospace and defense? And what are you seeing in the demand backdrop there?

Certainly. So in aerospace and defense, we have 3 businesses. We have business in aviation. We have business in water machines, offensive and defensive weapons and then we have business in space. They happen to be about 1/3, 1/3, 1/3. So aviation business is doing very good. Boeing wasn't building planes for a while when the MAX planes were down.

And now they're rebuilding, they got a 10-year backlog and those planes are loaded with our products and the demand is very, very strong. So that part of the business is doing very well. The offensive and defensive war machine, they had built this ammunition arsenal over 20 years, and they used it up in 2 months, half of it. And now they're trying to rebuild it.

So President Trump is asking prime suppliers that build these to ramp the production by 4x plus. I think the number I heard is 4x to 8x. Nobody can get to 4x, but they don't even know how to get to 4x. And for that now -- and I'm talking about primes like Raytheon and Knight Technologies and Honeywell and others. And so they're all ramping production. They're asking us.

I've spoken to CEOs of primes. They're asking our capability to ramp up. And we are in every single offensive and defensive weapon. We're in every missile, every plane, every radar installation, every drone, every interceptor, we're the largest supplier to that market, and we have broad penetration with all sorts of devices from diodes to controllers to voltage management to power management to RF to everything.

We are just huge. And so that business should do very, very well. But aerospace and defense business, it doesn't rise at the rate of the data center, but it has -- it prolongs. It will happen over multiple years. Some of these high reliability parts are -- have a long life cycle to build. It takes 9 months to build the parts sometimes.

So we're getting orders, backlog is building, and this should be a very, very good business in the coming year. It will last multiple years. The third piece of that is space. Now we are the largest supplier of rare hard or radiation hard parts to space. And space has 3 levels itself. There is a low orbit, then there's a medium orbit and there's deep space. And there's need a different level of radiation hardness to go to those 3 levels of space, and we make all of them.

And world has a new fascination with space. There was a huge fascination during Apollo times and all that. And then it kind of went away and we went to moon 40, 50 years ago, and then we didn't even go close. And now we want to land back on the moon in the next 2 or 3 years.

So a recent trip we had around the moon had millions of dollars of Microchip product on it. And as we are preparing to go to moon in 2028 and land on the moon, I think all the missions leading up to it will have a large amount of product in it. Plus we want to go to Mars, we want to land on Mars. Somebody wants to build a data center and moon. God love him, we'll have a lot of product.

I mean not very close, but closer to home on the LEO side. There's a lot of positivity now because of the SpaceX IPO. Is that something we should think about you guys participating and benefiting from given your heritage in radiation hardened products?

Or is it a different set of components than what goes further to the moon?

Some and some. I mean, Elon doesn't do anything in the normal way. So it's an incremental opportunity. But the difference is the regular space satellites that are launched by others historically are looking at longevity in a decade. It should last 20, 25 years.

Someday, slowly, it will come back in and burn off, but it will last long time. These constellation makers are looking at launching these things in volume with a high amount of redundancy, and they don't care if it burns up and comes back in, in 5 years. They'll just launch thousands more.

So they want to do it cheaply, but they want to do it in volume. So they're not using as much radiation harden devices. They are using many of the industrial and automotive-grade parts but using a large amount of redundancy to achieve the performance.

So therefore, the price level is a lot lower for the product. The business model is totally different. Business model looks a lot like the automotive industrial business model than the space business model. But the parts shipped there, if we know they're being bought by those companies, they would be an A&D business, and it's an incremental opportunity, but they also buy a lot of those parts from distributors.

So lower ASP, higher replacement cycle.

Yes. They buy those -- if they buy those parts from distributors, then we sometimes don't know who's buying it or what application they're going in to.

One of the markets we didn't talk about yet is automotive. I think you guys are optimistic that your portfolio of microcontrollers, Ethernet connectivity, 10Based-T are going to benefit from the shift to software-defined vehicle. How should we think about that in relation to your other growth drivers? And how much can that growth accelerate versus what it's done in the last few years for your auto business specifically?

So we're seeing a lot of auto growth right now, but they're not coming from T1S or ASA because those 2 protocols are still in design. Our growth is coming from automotive inventory running out, lots of new designs we have won in the regular USB, Ethernet, MOS bus and others.

So our automotive business is doing well, growing at a good rate. But what's layering on that is we're leaders in T1S and we're leaders in ASA. And essentially, what car manufacturers are trying to do is replace all the old protocols of CAN bus, LIN bus, MOS bus, Ethernet and everything else, RS-32 into a single T1S standard, which runs on a single twisted bear Ethernet wire.

So they don't have to use -- see, the past, all these modules were not connected. All these things didn't talk inside the car to each other. So the protocols could be different. When the car is getting all connected, software-defined, then they all have to talk to the central computer and talk to each other. And if they're not on the same protocol, then you have to build the bridge, the USB to Ethernet bridge, USB to MOS bridge, [indiscernible] network to Ethernet bridge. And these bridges are expensive and heavy and all that.

So therefore, there is a drive to take away all of those protocols and consolidate into a single standard. And we're leading that standard. We've got leadership devices in there. We have design wins with every major manufacturer. And these are '28, '29 production. It will largely begin in '28. Some could be late '27.

And they will add longevity to the cycle as the inventory-driven growth normalizes and we're getting growth in lots of other segments, and this will be just another layer. And when that begins, it will be a 10-year cycle where these protocols will merge over a long period of time.

Okay. Well, we're running out of time. So I'm going to ask Eric 2 questions at once. Eric, I think you're getting back up close to your 65% gross margin target. Inventory write-offs are pretty much done, but under loading charge is going to be here for a bit.

Can you talk about the levers to get back to that 65% level? And then also, can you speak to uses of cash right now? I know delevering is a priority, but how much longer do you think it will be until we can start to see some repurchases again also?

Yes. So the gross margin story is pretty simple at this point in time, right? If you take our gross margins from last quarter and add back $46.6 million of underutilization charges, we're at our long-term target. So it's a matter of growing back into our capacity. We're ramping fast.

I don't expect those underutilization charges to completely disappear by the end of the year, but they're going to be reduced significantly. So that progress is just blocking and tackling and growing back into our capacity. From a use of cash, we still are relatively highly levered. We do expect that our net leverage will fall below 3 this quarter, which is a good milestone for us.

But there's still work to go. We want to continue to reduce debt, make sure that our balance sheet is strong and can withstand the next cycle, which there will be another cycle in semiconductors. So we're going to continue to generate cash. Earnings are going to improve, EBITDA is growing, but the cash generation is going to go to support the current level of dividend and pay down debt.

Okay. All right. Well, gentlemen, we're out of time, but really appreciate you joining us. Thank you, Steve. Thank you, Eric.

All right. Thank you.

All right. Good morning, and again, welcome to JPMorgan's 54th Annual Technology Media Communications Conference. My name is Harlan Sur. I'm the semiconductor analyst here for the firm.

Very pleased to have Eric Bjornholt, Chief Financial Officer of Microchip here with us today. We also have Sajid Daudi, who's Head of IR. Microchip top 5 microcontroller supplier globally, #1 market share in industrial market MCUs. 60% of the MCU portfolio is advanced 32-bit in higher architectures, solid analog power management portfolio, #3 market share leader in FPGAs.

It's been our busy earnings season, so I've asked Eric to start us off with a summary of the March quarter, June quarter outlook, an update on the 9-point recovery plan that Steve laid out in March of last year, and then we can go ahead and kick off the Q&A. So gentlemen, thank you for joining us this morning. And Eric, I will turn it over to you.

All right. Hey everybody. Thanks, Harlan. Appreciate the introduction. So during the course of this discussion, we will be making certain future forward-looking statements about the financial performance of Microchip, and I refer you to our filings with the SEC that identify important risk factors about the company. So a quick summary of our March quarter results. It was fiscal '26, which ended in March was a very good year for Microchip. March of '25 to March of '26, the company was up 35% in revenue our EPS went from $0.11 to $0.57 on a quarterly basis. So that definitely turned things around.

We made significant improvements in reducing the inventory overhang that we had. So inventory came down by I think $320 million, something like that from December of '24 to March of '26. So great progress there. We are ramping our factories as we speak because we don't want inventory to go too low. I think distribution inventory was corrected during that time period. Customer inventory is in a much better spot than what it was. And Steve Sanghi kicked off this 9-point plan in March of the prior year, and we've really executed well on that. We adjusted our manufacturing footprint. We shut down a factory, 1 of our wafer fabs in Arizona, and we are rightsized today with plenty of capacity to grow into.

Like I said, inventory has come down. Cash flow has improved significantly. Our leverage, which really went quite high during the down cycle is now back to a much better spot. I think we should end the June quarter with net leverage below 3, which is a good milestone for us, and we'll continue to make great progress from there as we see continued significant increases in EBITDA. So we're extremely focused today on getting towards our long-term operating model, which is a 65% gross margin and 40% operating margin.

Our guidance for the current quarter is for revenue to grow about 11% sequentially. The gross margin target is 62.5%. So you can see that we're making good progress towards that 65% target our operating expenses. We still have some work to do on that. The forecast for the current quarter is about 29% versus our 25% target, but overall operating margins are trending nicely towards the 40%, and we see a clear path to get there.

So with that, I'll turn it over to Harlan for questions.

Thank you very much. That was a great opening commentary. I appreciate having the team here. I'm going to start with some of the near-term dynamics in fundamentals. So you guys -- you called the bottom of your business in the March quarter of last year when bookings inflected meaningfully in that quarter. And since then, book-to-bill has remained strongly greater than 1, backlog has continued to grow.

First phase of the industry recovery is always excess inventory is bottoming at your customers, chip orders and proving to the level of your customers' consumption trends, right? Second phase of the recovery is restocking by customers in anticipation of continued sort of strong demand. You're starting to see restocking activity amongst your distribution customers. But are you also starting to see restocking activity in your direct customer base as well?

So it's harder to track, right? In distribution, we get inventory reporting from the distributors each month. So we can -- we know what their inventory is doing and their inventory is at levels now that are lower than they probably should be at about 26 days, and we think the restocking effort is happening there. And that word alone sometimes scares investors. But when we are projecting growth of the business at 11% sequentially, distributors need to restock, they need to build inventory just to keep up, so their months of inventory don't decline again. So I think that's normal behavior.

And direct customers, I think it's a mix. I think we still have some customers that are likely still working through inventory, but many of them that we haven't seen purchase from us for 12 or 18 months are coming back. And so I think inventory is getting rightsized there. It's just a little bit harder to measure.

Okay. Got it. Asia region, this is kind of interesting, which includes China came back very strong in the March quarter. It was the fastest-growing region in your March quarter, both sequentially and year-over-year, not sort of what we would expect, right, given Chinese New Year's holiday season, can you talk about the demand trends post Chinese New Year's and was the profile of demand out of that region fairly broad-based? And maybe what are you seeing so far this quarter out of the China region, Asia region?

Yes. So the recovery that we saw in Asia and in China was broad based. It was really across all of our end markets, nothing particular to call out there. Many of the distributors that we sell through in that region are thinly capitalized. And because of that, they really took inventory down hard and are now at a point where they have to be buying at least in line with where end consumption is. So I think that's probably the biggest driver, but it's broad-based.

And typically, post Chinese New Year's, obviously, given that the March quarter was strong. I mean, it seems like post Chinese New Year, you guys did see a pretty strong snapback in orders as the workforce came back to -- post the holidays. It sounds like sell-through was strong, and therefore, that led to sort of follow through and orders out of that region post-Chinese New Years.

That's correct. So sell-through activity was strong. We saw a nice bump in sell-through really in every geography last quarter, and that's continuing into this quarter. bookings activity was quite strong. We had a very strong book-to-bill in the March quarter. We highlighted in our earnings call that April bookings were the highest month of bookings that we've seen in almost 4 years.

So that trend has continued and is continuing into May. Bookings have continued to be quite strong, and it's broad-based again. I can't call out one particular end market or one geography. It's coming back across the board.

Good to see that trend continuing. On pricing, it's clear that many of your peers are starting to take up pricing in some beginning of this year, some in the second half of this year, a combination of strong demand, but rising input costs. What I said, I think Steve talked about monitoring input costs, looking at segments where product availability is tight customer by customer sort of basis.

But at this point, the team has not raised pricing. As you look into the second half of this year and next year, and the potential strong demand profile in front of you and your customers, is the team considering price increases?

So we are not considering any sort of broad-based price increase at this point in time. We've got a clear path to get to our gross margin target with where the cost structure is today. So I'm not going to sit here and say, "Hey, if there are significant supply chain cost increases, we will have to do something to pass those costs on". But we've been in a mode of wanting to regain market share. We wanted to rebuild customer relationships after some of the missteps we took in the last cycle. And we are seeing strong evidence of that with how our customers are coming back, opening up new designs for us. We want to keep that momentum going.

But there's no doubt if supply chain costs continue to rise, that we will have to do something. And we want to be very fair with our customers, at least our partner customers that have opened up all these design opportunities for us that we don't do anything to negatively impact that relationship. So it's a balance, and we really want to maximize our operating margin dollars long term. And if we can gain share and keep prices relatively flat, we will do that.

The one area where pricing is going up is as your customers build their systems, looking at their building materials, memory pricing for them continues to increase very meaningfully, right? And we do see some pullbacks in production on low to mid-end smartphones, PCs, consumer electronics, those segments of the market.

But what we've heard is even like big-ticket items like consumer appliances and industrial applications are being somewhat impacted by this dynamic, right? And I know consumer and PC is a small part of the team's overall business. But is the team seeing some of this memory dynamic being reflected across parts of your customer base?

In other words, maybe not as we look at the more seasonal second half of the year, your customers typically would be building in anticipation of that. But maybe for some of your customers because of some of the higher memory prices, maybe unit volume builds are a little bit lower than maybe what they've done historically. Have you seen any indications of memory pricing increases on customer [ new ] plans?

We've seen a little bit of that, right? So when something is constrained, customers can't necessarily build everything that they want to build. And with that, they might not be able to complete their system and because of that want to have us hold back on delivering to something them until a later date when they think they can get the entire bill of materials to build their system.

Specifically for our business, we do have a serial E-squared memory business. It's a small piece. We reported in the other category of revenue. It's not part of the microcontroller and analog revenue. And we have seen that business pick up as others are reallocating their capacity elsewhere. We're gaining some share there and seeing good bookings and order momentum for that piece of the business, but it's small.

Got it. Have you -- could you give us a sense like -- I've been asked this question before, my response is like it's like single digits percentage of Microchip's revenues. But if I were to look at the PC, smartphone, sort of consumer electronics, end market segments, do you have a sense of the rough contribution to your revenues for those particular end market segments? I mean, it's small. I just...

Yes. It's quite small. I mean we break out consumer as 8% to 9% of revenue overall. So it's really our smallest market segment.

Got it. That's -- actually, before I turn to the mega trends, let's see if there's any questions from the audience. If you do have a question, raise your hand, and we'll get the microphone over to you. Any questions?

Let's talk about the technology, the portfolio, the mega trends and data center and AI. The team has done a great job of helping us to understand your strong position, enterprise SSD controllers, PCIe switching, signal conditioning, networking, timing and security, right? This is a subsegment of your data center and compute end market, which was 18% of your overall revenues last fiscal year.

Most of your peers in MCUs and analog are driving their data center businesses currently like anywhere from 50% to 100% year-over-year. Any rough quantification on the current growth profile of your AI-focused products and maybe the percentage of your data center and compute business that represents AI?

Yes. So we haven't broken that out, and we're getting a lot of investor questions on it. So we are working on a way to break that out separately. So included in our end market that we break out as data center and compute, those are standard compute, which isn't really in that category. We sell a lot of kind of standard microcontroller or timing products, security products into that segment, too, that aren't going to exhibit the type of growth that you're talking about, 50% annual growth.

But what we've been trying to do is highlight our Data Center Solutions business unit. We've had Brian McCarson, who leads that business unit on 2 of our last 3 earnings calls, highlighting some of our product initiatives there like our PCIe Gen 6 that we brought to market, that's really the only product in that market that's on 3-nanometer from TSMC and bring some significant power savings to our customers, and we're seeing good momentum there.

It's not shipping in volume production today that starts to happen near the end of this fiscal year and more significantly into next fiscal year, but the momentum is behind it there. And we will find a way to provide investors better insight into that. And you can see that, that piece of the business is growing at the rates that you're talking about.

Perfect. On your total system solution strategy, this has been a dynamic and an initiative that the team has been talking about and focused on for quite some time, right? And the idea is to drive higher attach rates of Microchip's products to your anchor products, those anchor products being microcontrollers, your microprocessors, you have FPGA products, more analog attached, more power management attached, networking chip attach per MCU/MPU design in, right?

A year ago, you presented an insightful -- I thought it was a pretty insightful chart on anchor products, showing the attach rate that drove an average of 4 to 5 additional Microchip products per anchor product design win, right? And as you look at your design win pipeline over the past 12 months, I mean, has that attach rate remained at these levels or increasing? And what programs has the team put in place, sales, marketing, channel partners, reference platforms to drive the attach rates even higher going forward.

So I think the attach rates are likely similar to that growing modestly from there. That's pretty decent attach rate. But what we're continuing to do is roll out additional reference design with these anchor products, educate our sales team to take those to the customer. And really, we want to gain more and more share of the bill of materials out of customer by using this TSS strategy.

So it's working quite well. It's fully ingrained in our business units working together whether it's complementary products that sell together to bring those to market effectively and make the customers' job easier as they design their system and have reference designs that are working parts that work together easily and speeds their time to market. That's really important from an engineer's standpoint.

And on the team's 6 megatrend focus, Edge and IoT compute, data center AI, sustainability, e-mobility and networking, right? From fiscal '21 to fiscal '24, the megatrend revenues grew at a 26% CAGR, right? That's 2x faster than the overall franchise and represented 47% of your total revenues in fiscal '24.

I know in calendar '24 and part of calendar '25, the team was still working with customers on inventory work-downs. But since the recovery profile started in the first half of last year and into the first half of this year, have the megatrend focus areas started to outperform again, the broad-based business and any way to quantify that?

So the mega trends are all performing well, and it's still a significant focus with within our company today, mega trends are now more than 50% of our revenue, right? And to think that, that can continue to grow at 2x the overall rate in the company is probably not possible. But those areas are still high growth, significant focus rate that we have from a customer engagement perspective and a product development standpoint. So it's going extremely well. I think that strategy has played out as we expected and continues to be high growth for the company.

One of the other leadership areas that Microchip participates in is aerospace and defense. And as that spending continues strong, a big part of that is your leadership in Field-Programmable Gate Arrays what we call FPGA products, right? The market share rankings out are out. Your FPGA business continues to be a strong franchise. #3 global market share leader, $0.5 billion per year franchise. This business outperformed your overall business in calendar '25, you've got your PolarFire product line, which is a leader in aerospace and defense applications.

Outside of aerospace and defense though, we've seen a very big pickup in demand for mid-range FPGA solutions across a variety of applications. It's a highly gross margin accretive product family, PolarFire, your next-generation PolarFire 2 family. How is the team doing on expanding your leadership from aerospace and defense into that sort of broad market sort of midrange side of the market.

Yes, it's a good question. So FPGA is an important area of focus for the company. As you say, we're the #3 player there based on data that's recently been released. So the company has been really focused on areas outside of A&D that has really driven that market leadership and allowed us to gain market share really since 2013. We bought a company called Microsemi in 2018. They had bought a company called Actel for these products originated from.

And with PolarFire, with the upcoming introduction of PolarFire 2, we have really great capabilities across the midrange and have areas where we can go into the low-end FPGA and the higher-end FPGAs. We are really known for our power efficiency, physical security, and really the benefits that we bring to anything that needs like radiation hardened products, which is really helpful for us in A&D, which is a fast-growing market at this point in time with a lot of the replenishment that needs to happen and the arsenals that have been used around the world on the defense side. I know your question was really more focused on other areas like industrial, where we've expanded, but the company is doing well. We're continuing to double down on our investment there. And I think the opportunities are large for us going forward.

Perfect. Before I go...

Go ahead. Go ahead, Sajid.

I'm just going to add to it just a little bit that I think as you kind of think beyond just the aerospace and defense on the industrial complex, as you think about vision capture, smart agriculture. So the use case is really expanding as you think about Industry 4.0 type of applications. So FPGA kind of plays a big role there as well. Still early in its infancy stage, but certainly starting to pick up momentum.

Before I move on to -- I want to talk about manufacturing footprint and future expansion plans. Is there any questions from the audience? Again, if you have a question, just raise your hand, and we'll get a mic over to you.

Your wafer outlet mix is 35% internal, 65% external even after the Fab consolidation. What's the longer-term mix target associated with your new financial targets you outlined that in your prepared remarks, right? 65% gross margin, 45% up margin targets, revenue growth that industry plus. So what is embedded within that? How do you sort of that longer-term goal? How do you think about your wafer output mix internal versus external?

So I think generally, it's going to stay roughly in that 35% range, plus or minus for the coming years. Now some of our faster-growing areas when you think about the data center product and FPGA and some of the networking products, they are outsourced to the external foundries. And so maybe there's a small downtick that comes from our internal factories as a percentage of the total as we see some higher growth coming from some of these exciting areas that we've talked about.

But there's really nothing that's changing from how we're running the business. Really, anything that is 110-nanometer above can be manufactured internally. And I think 90-nanometer or below, we use an external foundry partner for. And then on the assembly and test side, we do about 70% of that in-house, either in Thailand or Philippines. And when we make the decision to bring something in-house is when the volume is high enough to justify the investment. And right now, that's 70% of the total. But I expect over time that, that percentage will inch up modestly.

Got it. Your Fab 5 in Colorado is still 6-inch diameter wafers. So do you have the capability to expand your output without any new greenfield expansion? And as we look forward in time, as your demand profile continues to improve, is this an option to the Microchip team as your business continues to scale higher, upgrading Fab 5 to 8-inch wafer diameter over the next few years. Is that still an option to the team?

It is still an option. So we shut down our Arizona fab. Fab 2, which was an 8-inch Fab. We expanded clean room space in both Oregon and Colorado. Oregon is our larger factory in...

That's 8-inch, right?

That is 8-inch. That is 8--inch. And Fab 5 does have the capabilities to do 8-inch longer term. We have not made that investment as of yet, but we could in the future without having to expand and build additional clean room space.

Got it. The team does have -- and we haven't revisited this in a while, but the team does have a China for China strategy, right, to serve your larger base of domestic China customers. The plan, I think, was to partner up with a lot of the China-based test and OSAT guys out there and ship your chips or dies to China. Your China-based OSAT partners will do the packaging, final test and then fulfillment, does the team already have that strategy in place? Or is that something that's sort of still on the come?

So we really abandoned that strategy. With what's happened on the tariff side and China really looking at the point of Fab being what is declared from a tariff perspective, there wasn't a big benefit for us to do that. And so really, what we're focusing on is continuing to bring cost-effective solutions that customers want to use to the marketplace. Obviously, our China customer base has chosen Microchip over the competition over time for those reasons.

Now we know that there's clearly going to be demand, particularly for domestic-based China companies to buy locally. And we can only offset by bringing products that, that competition doesn't have to them and doing it in a way where they are incentivized to use Microchip because we have the best product at the right cost structure for them to use. So that's really the focus.

And you had talked about 70% of your assembly and test being, I think, done in-house, right? And part of the 9-point plan was to take your Philippines and Thailand, complexes. You were going to optimize those footprints as well. How did -- did -- are you already through that sort of optimization process. And so without the China for China plan, I'm assuming that you're leveraging Philippines and Thailand for packaging and shipping across all of China -- I mean, across all of Asia, including China, is that kind of the right way to think about it?

You are correct. Really, the optimization that was -- that happened there was just head count reduction, which was mostly attrition because there's higher attrition that happens in those locations. So we are ramping head count in all those locations, particularly in Thailand today. Those are larger operations from an assembly and test perspective, there's a great workforce there to tap into, and we're being able to ramp quite efficiently. So we're in a good spot in ramping assembly and test operations.

Now that inventory write-downs have normalized. You still have about, I think, as of last quarter, $47 million per quarter of underutilization charges, 3.5 percentage points of gross margin goodness, right? This alone would get you to your target 65% gross margins. Given the significant acceleration of the demand profile that the team has seen very broad-based, is it possible that the team can wind down most of these charges exiting this year?

So my anticipation is that those charges will come down significantly, but won't be gone by the end of the fiscal year. So we have some limitations on how fast we can ramp our internal factories. We can only ramp our internal wafer fabs by like 15% to 20% in a quarter. And today, they are producing at a level that is well below than what we're shipping, and that's why inventory has been coming down.

So we are moving as fast as we can to add production specialists or direct labor force to particularly the Oregon factory to ramp it as quickly as we can. But we will still have underutilization charges when we leave the fiscal year, but they will come down each quarter. I think last quarter, it was down $6 million or something like that. It will probably come down a little bit more than that this quarter as we're really in heavy ramp phase, but it takes time.

Got it. There was one product question that I forgot to ask. But before that, any questions from the audience? On the product side, obviously, a top 5 supplier global microcontroller sort of market share, you actually do have a very diversified portfolio into the embedded processor space. You've got your 32-bit microprocessor. You have a hybrid microprocessor family. And in 2024, you introduced your next-generation very high-performance 64-bit embedded processor family for edge compute applications, where we are seeing a lot of activity.

Whereas your MCU and 32-bit product lines were ARM-based and PIC architecture based, the team chose to enter the 64-bit microprocessor arena with a RISC-V based architecture. This embedded processor market is a $5 billion market opportunity. Can you just give us an update on the team's efforts to penetrate the embedded processor market? And will the team rollout a more industry standard sort of ARM-based 64-bit processor family of products?

Yes. So we made the decision on RISC-V based on back from the customers that we were targeting with that particular product set. I believe that over time, we will have a mix of both RISC-V and ARM-based products in the 64-bit arena. And really, those decisions will be -- we'll choose the architecture based on the functionality of the product that is needed and which architecture suits at best.

So we've tended to be architecture-agnostic over time. We've had a bunch of different architectures in the portfolio and then make that choice based on the product, what the requirements of the customers are going to be. But it is a big market opportunity for us. You mentioned it's a $5 billion TAM opportunity, and we're continuing to introduce new products that we are doing quite well in the early stages of introducing products, both in 64-bit MPU and then MCUs are coming also. So the team is tracking well on that trajectory.

Sajid, anything that you would want to add on the high-end compute?

No. I think other than just that's maybe just a little milestone moment for us where recently, NASA and JPL kind of put out a press release on a product that we collaborated with them, which the computation and 500x what was set out to be originally than what the products are available. So again, small thing, but as I said, in the compute family, that was a good milestone for us.

Yes, that's the high-performance space compute.

Got it. On the financials, as I think about the financial model, your focus and things like you're focused on the 6 megatrends, your product evolution that includes augmenting the portfolio with high-performance compute networking, storage products, our performance in aerospace and defense, really good traction in the data center segments of the market.

It would seem to me that the product gross margin should be mixing higher going forward. In other words, as you look at your design win pipeline, you look at your development pipeline, your design win pipeline, over time, it feels like your mix is getting richer and richer and richer, right? Any way to quantify your design win pipeline and the product gross margin profile improvements you anticipate over the next several years as you create these more kind of higher value-added solutions for your customers?

Yes, it's a good question. So in some of these faster-growing areas, whether it's the data center products that we've talked about, our networking and connectivity products, FPGAs, these product lines tend to have higher than corporate average gross margins. And we think the growth from those product areas is also going to be higher over time. So I think we have a positive mix shift there that's going to be going to benefit us.

Obviously, we're going to continue to be competitive, trying to win share and price is a factor in that. But overall, we're very comfortable in getting to the 65%, you take away the underutilization charges and we're there. And then it's just a matter of how the business mix develops over time. And if we have good success in what we think are going to be our higher growth areas, margins could be in a very good spot.

We're not willing at this point to update our long-term margin profile. We are really targeting something that we can float around a 65% when the cycle is good and when the cycle is bad. And obviously, we need to do a much better job of managing inventory in the current cycle than we did in the last cycle for that to happen. So we don't have these large underutilization charges that we don't over invest from a capacity standpoint. But the team is focused on it, and I think our margins are going to continue to improve nicely.

Perfect. Eric, Sajid, thank you very much for your participation today. I look forward to continuing to monitor the execution of the team as the year unfolds. Thanks very much.

Thanks, everybody.

Greetings, and welcome to the Microchip's Q4 and Fiscal Year '26 Financial Results Conference Call. [Operator Instructions] As a reminder, this conference is being recorded.

I would now like to turn the conference over to your host, Eric Bjornholt, Chief Financial Officer. You may begin.

Thanks, Kate, and good afternoon, everyone. During the course of this conference call, we will be making projections and other forward-looking statements regarding future events or the future financial performance of the company. We wish to caution you that such statements are predictions and that actual events or results may differ materially. We refer you to our press release of today as well as our recent filings with the SEC that identify important risk factors that may impact Microchip's business and results of operations.

In attendance with me today are Steve Sanghi, Microchip's President and CEO; Rich Simoncic, Microchip's COO; Brian McCarson, VP of Microchip's Data Center Solutions business unit; and Sajid Daudi, Microchip's Head of Investor Relations.

I will comment on our fourth quarter and full fiscal year 2026 financial performance. Brian will provide an update on our Data Center business, and then Steve will provide commentary on our results and discuss the current business environment as well as our guidance. We will then be available to respond to specific investor and analyst questions.

We are including information in our press release and this conference call on various GAAP and non-GAAP measures. We have posted a full GAAP to non-GAAP reconciliation on the Investor Relations page of our website at www.microchip.com, and included reconciliation information in our earnings press release, which we believe you will find useful when comparing our GAAP and non-GAAP results. We have also posted a summary of our outstanding debt and our leverage metrics on our website.

I will now go through some of the operating results, including net sales, gross margin and operating expenses. Other than net sales, I will be referring to these results on a non-GAAP basis, which is based on expenses prior to the effects of acquisition activities, share-based compensation and certain other adjustments as described in our earnings press release and in the reconciliations on our website.

Net sales in the March quarter were $1.311 billion, which was up 10.6% sequentially, and up 35.1% over the year ago March quarter. Our revenue results were above the high end of the guidance range we provided on February 5, 2026. We have posted a summary of our net sales by product line and geography as well as our fiscal year 2026 revenue by end market on our website for your reference.

Our end market mix did not change materially in fiscal year 2026 compared to fiscal year 2025. Industrial was 31% of sales. Data Center and Compute was 18%. Automotive was 17%. Aerospace & Defense was 16%. Communication was 9% and Consumer was 9%. These percentages are our best estimates of the end market splits and there's probably a couple of percent [ error band ] due to the fact that almost 50% of our business and the long tail of customers are serviced through distribution, which makes it difficult to track the end markets.

On a non-GAAP basis, gross margins were 61.6% in the March quarter, including capacity underutilization charges of $46.6 million. Operating expenses were at 31% of sales and operating income was 30.6% of sales. Non-GAAP net income was $327.3 million, and non-GAAP earnings per diluted share was $0.57, which was $0.07 above the midpoint of our guidance.

On a GAAP basis, in the March quarter, gross margins were 61%. Total operating expenses were $582.2 million and included acquisition intangible amortization of $107.8 million, special charges of $6.4 million which was primarily driven by costs associated with the closure of Fab 2. Share-based compensation of $59.9 million and $1 million of other expenses. The GAAP net income attributable to common shareholders was $116.4 million or $0.21 per share. For fiscal year 2026, net sales were $4.713 billion, and were up 7.1% from net sales in fiscal year 2025.

On a non-GAAP basis, gross margins were 58.5%. Operating expenses were 32.2% of sales, and operating income was 26.3% of sales. Non-GAAP net income was $933.9 million and EPS was $1.64 per diluted share.

On a GAAP basis, gross margins were 57.7%, operating expenses were 47.3% of sales and operating income was 10.4% of sales. The GAAP net income attributable to common shareholders was $118.8 million.

Our non-GAAP cash tax rate was 5.8% in the March quarter and 8.6% for fiscal year 2026. The cash taxes remitted in Q4 were lower than originally forecasted, while our pretax profit was much stronger than forecasted, driving the March quarter rate down. Our non-GAAP tax rate for fiscal year 2027 is expected to be about 10%, which is exclusive of any tax audit settlements related to taxes accrued in prior fiscal years.

Our inventory balance at March 31, 2026, was $1.035 billion and down $22.3 million from the balance at December 31, 2025. We had 185 days of inventory at the end of the March quarter, which was down 16 days from the prior quarter's level, driven by our inventory reduction actions and increased revenue. Included in our March ending inventory was 15 days of long life cycle, high-margin products whose manufacturing capacity has been end of life by our supply chain partners. Inventory at our distributors in the March quarter was at 26 days, which was down 2 days from the prior quarter's level and at the lower end of what we have experienced historically. We expect distribution restocking to occur in the near term as distributors will likely grow their inventory holdings above current levels to support growth.

Distribution sell-through increased by 11.4% during the quarter and distribution sell-in was just modestly lower than distribution sell-through. Our cash flow from operating activities was $257 million, and our adjusted free cash flow was $228 million in the March quarter. As of March 31, our consolidated cash and total investment position was $240.3 million. Our total debt increased by $143 million in the March quarter, the increase in debt was impacted by our refinancing activities in the quarter, which included issuing a 0% 4-year convertible bond for which we paid $68 million for 100% capped call to provide some protection from future equity dilution from stock price appreciation.

Our adjusted EBITDA in the March quarter was $466.8 million and 35.6% of net sales. Our quarterly adjusted EBITDA was up 132.9% from the March 2025 quarter. Our trailing 12-month adjusted EBITDA was $1.496 billion, our net debt to adjusted EBITDA was 3.54 at March 31, 2026, and down from 4.18 at December 31, 2025. We expect the June 2026 quarter to be an excellent cash generation quarter for us, resulting in meaningful debt reduction and also expect our net debt to adjusted EBITDA to drop below 3.

Capital expenditures were $14.2 million in the March quarter and $91.1 million for fiscal year 2026. Our expectation for capital expenditures for fiscal year 2027 is to be approximately $100 million. Depreciation expense in the March quarter was $38.7 million.

I will now turn it over to Brian, who will provide some exciting insights into our Data Center Solutions business unit. Brian?

Thank you, Eric, and good afternoon, everyone. We are seeing significant momentum across all 3 major product families within our data center solutions business, and I'll summarize the progress we're making in each of them. Importantly, these wins are translating into higher content per system and a longer runway of production ramps, which we expect to support durable growth as the data center architectures continue to scale.

First, our storage controller products have supported some of the world's most reliable SAS, SATA, NVMe and RAID infrastructure over the past decade as a leading provider of data center storage control solutions.

And as AI inference and agentic AI workloads increased demand for persistent data access, demand for our products continues to grow. We've been strengthening our product road map and customers are responding positively. Most recently, our Adaptec, SmartRAID, NVMe storage accelerator received the Nimbus Innovation Award with benchmark results showing up to a 3x improvement in read and write performance versus a leading competitor's offering. This can translate into better XPU utilization in real-world data center workloads.

Second, our memory controller product family has been reinvigorated by the recent launch of our next-generation devices. We brought 3 new CXL and PCIe-based devices into production in calendar year 2025, and our next Gen 5 dual port device is scheduled to enter production this quarter. We've already secured meaningful design wins that have begun ramping and that we expect will continue to grow through fiscal '28. These next-generation devices have been externally benchmarked and our customers are reporting industry-leading performance in jitter tolerance, which is a key measure of storage controller performance.

Third, our Switchtec business has continued to build momentum since the announcement of our latest PCIe Gen 6 switch just 2 quarters ago. Since that time, we've secured a total of 6 significant design wins with customers citing our product quality, our signal integrity, differentiated features and our strong performance per watt as industry-leading. This is helping Microchip maintain its position as a leading PCIe data center switch provider, while gaining share in both scale-up and scale-out market segments. Our Gen 6 switch is scheduled to begin production ramp at the end of this quarter and many additional design wins are expected over the next year.

In addition, we're very excited to announce we've entered the PCIe retimer market this quarter. This retimer is architected as a companion die for our PCIe Gen 6 switches and is also designed to support the rapid growth of the active electrical cable market with compatibility spanning Gen 1 through Gen 6 speeds. Customer feedback has been encouraging and we have already secured a major OEM design win on an upcoming Gen 6 platform, displacing one of our competitors. Customers have been explicitly requesting that Microchip offer a companion retimer because sourcing both the switch and the retimer from a single vendor reduces implementation complexity and risk. This is helping drive these wins and strengthening our competitive position in new designs with an even more complete PCIe connectivity platform.

The wins I described across all 3 data center product families reflect several competitive advantages. First, power efficiency. Our Gen 6 switches and retimers deliver strong power efficiency relative to alternatives, which directly benefits operating costs and data center environments.

Second is feature completeness. By offering both the switch and the companion retimer, we provide customers with a more complete scale-up and scale-out PCIe solution, reducing the need to integrate products from multiple vendors. Third is quality and tools. We deliver the reliability, the performance and stability data center customers demand, while offering world-class diagnostic and configuration tools with ChipLink.

And finally, support through the entire customer journey. From initial design through production ramp and beyond, customers value a partner that has invested in their success at every stage. That combination of power efficiency, feature depth, quality and long-term partnership is helping drive these wins. We believe these factors position us well as design wins continue to convert into production ramps.

I'll pause here and turn the call over to Steve to provide an update on our business and the guidance going forward. Steve?

Thank you, Brian, and good afternoon, everyone. I will start by providing you a brief update on our nine-point recovery plan.

The first item was to rightsize our manufacturing footprint. This was completed, the remaining item left is selling our Tempe fab. We have several interested parties, but the deal has not closed yet.

The second was to bring the inventory down. We have brought the inventory down from 266 days at the end of December 2024 to 185 days at the end of March 2026. Our overall dollar value of inventory has decreased by $319 million from the December 2024 peak inventory of $1,356 million to March 2026 ending inventory of $1,037 million. We're now in a significant revenue growth mode. We expect our inventory will come down naturally towards our goal of 130 to 150 days as revenue grows, and we appropriately manage our manufacturing and foundry resources.

Third was the megatrend alignment. This goal was completed by creating a megatrend for AI replacing 5G and creating another for network and connectivity, replacing ADAS, which stands for Advanced Driver Assistance Systems.

The fourth item was business unit alignment. This goal was completed by realigning business units. We used to have essentially 2 pillars: microcontrollers and analog. We have now created 5 pillars which are microcontrollers, analog, networking and connectivity, high-performance compute, and artificial intelligence on the edge. We know a few requests for us to break out the revenue for some of these growth segments. We are thinking through this as some of the revenues intertwined and challenging to break out we will provide more visibility as we complete our analysis.

The fifth was to look at our distribution programs. This goal was completed by realigning our distribution program and how we compensate our distributors for various levels of demand creation and fulfillment. We also considered our network -- we also consolidated our network by terminating several small underperforming distributors.

Number six, was customer relationship improvement. We undertook a major effort to improve our relationships with a large number of customers where the relationship had deteriorated during the COVID cycle. We now consider our relationship with our customers to be good, but it is still an ongoing process.

Number seven, was the new business model. We unveiled a new business model in March of last year. The model contained a long-term non-GAAP targets of 65% gross margin, 25% operating expense and 40% operating margin. At the bottom of this last down cycle in March of 2025, we had a non-GAAP gross margin of 52% which has now improved to 61.6%. At the bottom of the cycle, we had a non-GAAP operating expense of 38%, which has now improved to 31%. And at the bottom of the cycle, we had a non-GAAP operating profit of 14%, which has now more than doubled to 30.6%. The improvements are ongoing towards our long-term business model.

Number eight, was the operating expense percentage. This was simply putting a plan together to achieve our long-term non-GAAP operating expense target of 25%. We have come down from 38% to 31%. We believe that revenue growth and productivity improvements will get us the rest of the way.

And the ninth, final one was CHIPS Act. We are essentially on hold given we are still growing into our current capacity.

So this was a comprehensive update on our nine-point plan. The only main items that remain are further reduction of inventory and making further improvements in our gross margin, operating expense percentage and operating profit percentage. The nine-point plan has been an extremely successful program that was very well executed.

Next, I will talk about our business environment. We are seeing recovery in all of our end markets: automotive, industrial, communication, data center, aerospace and defense, and consumer are all looking better. The strongest sales performance last quarter was in aerospace and defense sector. From a business unit perspective, the strongest performance was in FPGA products.

We believe that we have completed the distribution inventory correction. Our overall distribution inventory is now somewhat lower than normal. The distribution sell-in and sell-through was close to even last quarter. We are starting to see large orders from distribution, which show some restocking happening in the distribution channel. The distributors' customers' inventory has also come down significantly. After using up the excess inventory, we are seeing thousands of customers reengage in buying our products. Our customer count has started to go up. We're also seeing customers from new designs and from an improved relationships, start buying our products, adding to our bookings, revenue and customer count.

Now let's get into guidance for the June quarter. In the June quarter, we expect strong growth from data center, A&D sector, industrial and automotive end markets. From a business unit perspective, we expect nearly all business units to participate in the growth, thus broadening out this recovery.

Our bookings for the March quarter were significantly higher than those for the December quarter. The book-to-bill ratio for the March quarter was well above 1, resulting in a much higher backlog entering the June quarter compared to when we entered the March quarter. And April was the largest booking month in almost 4 years.

A comment about lead times. While lead time for our products have been 4 to 8 weeks for some time, we are continuing to see lead times increase on many of our products. We're running into challenges on certain kinds of substrates and subcontracting capacity and also foundry constraints on multiple nodes. These challenges previously were isolated to very specific areas, but are now starting to spread more broadly. Our customer requests for expedited shipments have increased significantly from a couple of quarters ago, pointing to some customers' inventories running very low.

Taking all of these factors into account, we expect our net sales for the June quarter to be up 11% sequentially plus or minus 1%. This at the midpoint would be up 35.3% from the year ago quarter. We expect our non-GAAP gross margin to be between 62.25% and 63.25% of sales. We expect our non-GAAP operating expenses to be between 28.75% and 29.25% of sales. We expect our non-GAAP operating profit to be between 33% and 34.5% of sales, and we expect our non-GAAP diluted earnings per share to be between $0.67 and $0.71 per share.

With that, Kate, will you please poll for questions.

[Operator Instructions] Our first question comes from the line of Timothy Arcuri from UBS.

Steve, thanks for the update on the nine-point plan. Another thing you were going to try to give us was, you were thinking about what the pro forma growth rate was for the company. I know if you pro forma everything, the growth rate is not as high as it truly is because you got out of a lot of businesses. So I know that's something that you've been looking at. Do you have any update on that? What do you think the pro forma growth rate is for Microchip?

Timothy, thanks for your question. I think we are just going through a phenomenal growth right now and trying to put a longer-term growth rate in this environment, I think it's challenging. Any number I give you, the growth rate in fiscal '27 which started on April 1 here would be substantially higher than that growth rate. So I think we'll push that out further.

Okay. And then Eric, did you give us inventory charges, I didn't hear that?

So we didn't. As we went through the quarter, we kind of implied that as we got through the March quarter that our inventory reserve charges would normalize. And that is essentially what has happened in the March quarter. When you look at the new reserve charges that we've taken, the benefit that we're getting from previously written off inventory are kind of netting to a number that is in a range of what we would expect on a go-forward basis.

So I think we're in a good position there. When you look at our gross margins, the biggest thing keeping us away from our 65% target at this point in time is our underutilization charges, which were $46.6 million last quarter. And if you divide that by the revenue and kind of add that future benefit back to gross margin, we're essentially at the 65% target. So we're in good shape. Inventory reserve charges have normalized. It's not really a headwind to gross margin at this point in time, and we're glad to have that behind us.

Your next question comes from the line of Vivek Arya with Bank of America Securities.

First one is kind of more near term, March and June, both well above seasonal. What is driving that upside? Is it a specific end market? Is it more distribution? Is it more direct? And I think on the last call, you had mentioned that some direct customers or some sort of distribution are still burning through inventory, are they done? And as we look to September, that also is usually a seasonally stronger quarter. Consensus has your September up 5% to 6%. I don't know whether you would describe your September quarter visibility at this point?

So Vivek, the way I will describe is I think there are 3 things happening. One is we have been telling you for a long time that the distributors are burning their inventory, direct customers are burning their inventory and distributors' customers are burning inventory. And they were all in various innings, but everybody was pretty much in the ninth inning. And we have seen the distribution inventory now fully corrected actually to be below normal and are seeing some restocking orders from distribution as they prepare for the huge growth.

Distributor customers are also seeing -- I'm sorry, distributors are also seeing their customers come back in droves because they have completed their inventory. So distributors are placing large orders to serve their customers who have not been buying a lot of products in the last year because of inventory.

And the other factor is, as I talked about, we have improved our relationship with thousands of customers and in many, many of those cases, we had design wins with the customers, but we were not getting our share of the revenue. In some cases, they had duplicate designs, and they were buying somebody else's product for a different model. As we look at our overall customer count, it has increased by several thousand customers. So we're seeing the customers reengage as we have improved our relationship.

And finally, I think we are seeing a meaningful improvement in end markets. A&D is very strong. Data center is very strong. Industrial is very strong. Automotive is coming back. A lot of the automotive designs we have incubated in the last couple of years are going to production. So we are seeing the end market strengthening effect also. Combination of all those have been really well above seasonal growth for June quarter, which we hope will continue for some time.

I think the last question that Vivek asked was related to September and visibility, and I'll just comment on it real quickly. Our backlog is growing nicely. At this point in time, our September quarter backlog is higher than the June quarter backlog at the same point in time. Steve spoke to the booking strength that we saw in April. So things look good, but we're not willing to make a call yet on kind of percentage of revenue growth in September, but things look good at this time.

Understood. And for my follow-up, I think you described aerospace, defense and data center. And what is interesting, though, is that when I look at your change in end market mix year-on-year, it actually showed that aerospace and defense actually came down somewhat and so did data center. I appreciate that these are kind of approximate numbers.

But I would have thought that, that mix would grow. So can you help us understand, right, are we doing the right apples-to-apples comparison? Are there certain traditional, older things in the data center that we should not be focusing as much on? So why did those -- that mix come down in fiscal '26? And how do you see that mix evolving in fiscal '27?

I think if you look at the A&D sector, there was a huge increase in A&D sector percentage from 2024 to 2025. I think it used to be about 10% of our business and it went up to about 18%. So A&D did not drop during the post-COVID cycle as much, and that's why its percentage had increased. Now some of the others did better as they caught up. So percentage kind of plays games because it's a whole year number and data center and A&D has seen very significant growth and improvement in the last 6 months or so. But when you look at the overall number, that's where the math is. But they're still the stronger sector, if you look at the last couple of quarters. And especially A&D had grown huge the year before.

Yes. Maybe the other point on data center is our data center business really kind of bottomed out in the June quarter of last year and we've seen significant growth since then. And so we've been speaking about that each quarter, but each of these end markets had a bottoming at a different point in time.

Your next question comes from the line of Matthew Prisco with Cantor Fitzgerald.

I guess, first on the data center side. Can you help breaking down the sizing of that market today and perhaps the exposure across those 3 buckets that you talked about? And then you talked about the 6 new customer wins now for the PCIe Gen 6. You kind of size that opportunity in any particular wins in? I think you talked about $100 million win last time. So any other sizing you could offer?

Brian?

Yes. Thank you for the question, Matt. So a few things around your queries. One is we don't readily comment on the specific customers that we design with. We are, however, disrupting our competitors in this space, and we're seeing our design wins displace some of the other incumbents, especially in the Gen 6 area. Typically, in the product life cycle, it's uncommon to see a lot of design wins prior to production ramp. Yet our 6 design wins that we've already secured are prior to us doing our production volume release at the end of this quarter. So given that fact and that once we're in production, we expect to see a steady increase in design wins, we should see a ramp in revenue and then hitting our stride from a volume perspective in production in the next fiscal year.

All right. And then on the gross margin side, you guys have talked about kind of ramping utilization in the front and back end. So can you maybe help us understand where these levels stand today versus more optimal levels? And then given the seemingly accelerated utilization push, how should we be thinking about underutilization charges declining from here?

Okay. So we are ramping all of our large factories at this time. We've been in serious inventory reduction mode over the last 5 quarters and have made great progress there. And now we're ramping our fabs as an example, and there's some limitations on what we can do there, and we don't want inventory to get too low. So we're maxing that out. And so underutilization charges are going to come down nicely is our expectation in the June quarter, but they don't go away. That's going to take multiple quarters for that to happen, but each quarter, they will be reducing and that we're also ramping our back-end operations in Thailand and Philippines to support what we needed from a finished goods perspective.

And we actually do about 70% of our volume in-house there. Whereas in wafer foundry or in the fabs, we only do about 35% of that volume internally. So they're all ramping. We are fortunate that in the last cycle, we put a lot of capacity in place. So it's just a matter of adding people, having the raw materials to start in the factory, and we're in a really good position to be able to respond to upside demand.

Your next question comes from the line of Chris Caso with Wolfe Research.

So Steve, based on your earlier comments, do you expect that now we're fully through customers and distributors taking inventory down and we're now shipping in line with end demand? And I recognize that may not be the same answer for every market segment. Just interested in your view overall and by particular segments.

So there is nothing really ever absolute. If we have 110,000 customers, you can never say every customer's inventory is corrected. But both distribution, distribution customers and our OEM customers' inventory is broadly corrected and lots and lots of customers coming back and buying the product based on their run rate because they no longer can reduce the inventory further. Are there some customers where they may be still high on SKU here and there? Yes, and that will continue to correct as we go. But the strength of the business, you see it coming back is because of thousands of customers are returning and distributor is buying products for their customers as well as restocking because the inventory has gone too low.

Understood. The next question is with regard to pricing. And we've heard from some others in the space from distributors and such that at least in some areas, we're starting to see some signs that pricing is moving higher. In some cases, it's about the tight supply you spoke about, in some cases, it's about input costs coming up. Can you talk about what Microchip is seeing? And what you are planning on doing with pricing as we go into the second half of the year?

So first, I'll talk about our philosophy and then what we're going to do this time. Our normal philosophy is that we engage with customers at the design location at the time of design and give customer a price which he can count on almost through the life of the design and then comfortably design with us and then be able to buy the parts at that kind of price. That sort of has been the philosophy probably for 30 years.

The post-COVID cycle was very unique with -- there were a lot of input cost increases in Microchip as well as everybody else raised prices. And I think we've also talked about during that time we hurt our relationship with many of our customers, and we have worked very hard in the last year to reestablish good relationships with thousands of customers. So how we're dealing with it right now is I think our gross margins are doing very well and heading very well towards the longer-term target as this underutilization goes away in the coming quarters as we are ramping the factories. So we are trying very hard to really stay on the good side of the customers and not do a indiscriminate broad-based price increase, at least to our partners, strong partner customers.

And beyond that, it will be a customer-by-customer. We're looking at our input costs. We're looking at pricing. In many cases, pricing is adequate. In certain cases, if the price that we got was very aggressive when input costs have gone up, then we may adjust it. But as we speak right now, we have not increased our prices.

Your next question comes from the line of Joe Quatrochi with Wells Fargo.

Yes. I was wondering if you could give us any color on just -- you previously talked about kind of revenue contributions from the megatrends, but and it's kind of difficult to break each one down. But any sort of help on the growth of just the megatrends in general in fiscal '26?

Rich?

So as a percentage of overall revenue, megatrends have increased, but we've broken them down and changed them. So we didn't release a breakdown of that this year in line with -- we stuck with our market review. But we've left the megatrends outside of that purview. Overall, megatrends have been increased. Data centers increased, our communications business has increased, our ADAS business, our networking and connectivity business continues to grow because we lead in 10BASE-T1S and a number of Ethernet products. And so all of those businesses continue to grow very well.

And then as a follow-up, I think in the prepared remarks, you had talked about like actively expanding capital equipment in selective capacities. Any sort of color you could provide on just where those areas are that you're increasing capacity?

Actually, we didn't say we were increasing capacity. That was not in the prepared remarks. I think we said we expect CapEx this year to be about $100 million. And a large part of that is going to be just a maintenance kind of capital, we have sufficient in-house capacity. The areas where we would be adding capacity would be testing capacity for our data center business, for Gen 6 switch as it starts to ramp and retimers and all that. There could be some capacity increase for FPGA where we see significant growth. It could also be some in the Ethernet T1S area. But really, our capacity increase needs are going to be -- our capital needs are going to be modest because we have a lot of capacity.

Right. So we're really mostly just growing back into capacity that was put in place in prior years in the last up cycle.

[Operator Instructions] Our next question comes from the line of Vijay Rakesh with Mizuho Securities.

Just going back on the aerospace, defense side. As you look out, given you are one of the biggest suppliers there and given all the geopolitical tension, wouldn't you expect that business to start to accelerate into fiscal '27, '28 here? And I have a follow-up.

So aerospace and defense business is doing very well. We're getting large orders and it's a good business and it will grow but you also -- also think about for that business to grow significantly, our primes have to grow their capacity. So we can provide more parts for a missile, but the missile production capacity needs to increase. So we are right now talking to every major prime. We're trying to build new facilities to grow their production so it's going to take some time from that standpoint.

Secondly, the cycle time to produce parts for that sector are very long. Sometimes it can be as long as 9 months of cycle time because you build parts and then you park the lot and take a sample, burn them in for many months and if there are no failures, then you can ship the lot. So the cycle time in that sector are very long. So that's why it did very well in the last year. It's doing well now, but it doesn't really go up like a hockey stick that can happen in a cell phone market or data center market. It's very steady growth.

Got it. And back on the data center side, when you look at that portfolio of the PCIe Gen 6 switches and CXL which is probably starting to pick up here and storage controllers. How does that data center portfolio grow? Obviously, a lot of focus on data center and spend there. But any way to size the growth or the pickup there?

Brian?

Well, there's a few important factors that are happening right now in the overall data center ecosystem. We've been experiencing a significant super cycle with respect to AI-based data centers in recent years, but that's been primarily for AI training workloads. And we're now starting to see increases in Agentic AI and inference-based workloads. And those require a significantly larger amount of access to components that are traditionally PCIe based. So CPU utilization, NVMe and storage access is heavily impacted by inference and Agentic AI workloads.

So we believe we're positioned to grow well with the market in the coming years as we see that become more of the predominant growth area in the data center market in the future. And that's across our customer base, which includes hyperscalers, it includes OEMs and traditional enterprise server manufacturers as well.

Your next question comes from the line of Joshua Buchalter with TD Cowen.

Congratulations on the very solid results. Maybe following up on Chris' question on pricing from earlier. I mean, your message was very clear on rebuilding relationships with customers and not raising pricing. I guess how aggressive are you observing your competitors on pricing right now? And is it at the point where you're able to lean in and gain share as a result? And I guess, is there a certain level of input costs rising where you'd have to change its philosophy a little bit in the near to medium term and raise pricing as well?

So I mean, I think you're partially answering the question in your question. We are repairing our relationship with our customers, and many of those are repaired, and we don't want to stress them.

Number two, we are watching very carefully what every competitor is doing to the extent we have information from the market. And number three, we have strong partners, and we have people that have a tactical relationship, they just buy parts. And so we're separating those 2 in a way. Partners mean one thing, and the people who have just a casual relationship mean another thing. And if we see any stress on our input costs in a specific area, then we will be willing to adjust prices -- even non-partners.

Got it. Okay. You mentioned some growing areas of tightness in the supply chain. You made the decision a couple of years ago to not invest in a 300-millimeter fab and are investing in a more targeted way now. Given we're still not that far away in the grand scheme of things from the shortages a few years ago. I was hoping you could maybe reflect on your confidence that there's enough foundry capacity and investment going on right now that you and I guess the industry as a whole have enough room to run in what seems like a growing up cycle?

When we were willing to invest in 300-millimeter capacity, it was going to be under a license from a major partner. Without that license, we don't really have 300-millimeter IP and starting stand-alone, building a fab, developing a process and then developing thousands of mask sets on it is really just not a very cost-effective exercise. So we were doing it when a major foundry partner had told us that they will grow capacity largely on the bleeding edge. And therefore, the trailing edge or medium edge capacity will not grow and they were willing to give us a license. That thing is really no longer true.

In the post cycle, there was plenty of capacity until very recently, there was a plenty of capacity. It's becoming a little tight now, but the partners are adding capacity, and we believe we'll have sufficient capacity available in the coming years. And the major tightness would be really on the bleeding edge like 3-nanometer. The rest of it, we are getting what we need. Is it tight? Yes, but it's really not horribly bad.

Your next question comes from the line of Harlan Sur with JPMorgan.

Memory pricing continues to increase fairly meaningfully and we do see some pullbacks in production of low-end, mid-end smartphones, PCs, consumer electronics segments of the market. I know this is a smaller part of Microchip's revenue mix. But even some of the big ticket items like consumer appliances, industrial applications are being somewhat impacted by this dynamic. Like, for example, as we've heard of some of the industrial customers having to scramble for memory, not because of pricing, but because of end of life of older DDR4 memory and now they're having to kind of requalify their systems. Is the team seeing some of this memory pricing dynamic reflected across your customer base?

Yes. So you're right. I mean I think it fits into specific areas. So driven by AI, the DRAM capacity is very tight. So what some of the NAND flash manufacturers have done is shifted some of that capacity to DRAM. And then the lower end of the NAND flash capacity, some of the NOR flash manufacturers have moved into NAND flash.

So the NOR flash got constrained. So some of the E-square manufacturers in Asia shifted their E-square capacity to build a NOR flash. So everybody has moved upstream. And the bottom of that pile is E-square, which is what we make, Serial E-square and number of people have really basically abandoned that, double the price, which really means they have abandoned it because it converted the capacity to something else. And we are seeing significant opportunity in that business. It's not a very large business for us. But I mean, it's up significantly.

That's very, very specific memory-related but our run-of-the-mill microcontrollers, 8, 16, 32-bit microcontroller, analog, power management, A2D converters, D2A converters, our BiCMOS, DMOS other products, I mean they're all fine. Can we double any of those short term? No, capacity is tight, but we have sufficient capacity to meet the numbers we're giving you, even create the upside of the demand and backlog were to come in. And that capacity will continue to increase in the coming quarters. And the business really remains looking for turns. It's not like we're turning business away because we don't have capacity.

Got it. And as aerospace and defense continues strong, a big part of that is the leadership that you guys have in FPGA. I think the market share rankings for last year out here. FPGA business continues to be a strong franchise, #3 global market share leader. This business outperformed your overall business in calendar '25. Obviously, PolarFire is a leader in aerospace and defense applications. But outside of aerospace and defense, we're seeing that there's strong pickup for just general-purpose mid-range FPGA applications across the wide variety of end markets. It's a highly gross margin accretive product family, PolarFire, PolarFire 2. How is the team doing on expanding your leadership in midrange to broader market applications?

Well, we are. We absolutely are. The PolarFire 2 is in the fab and shortly coming out, and we're very excited to be launching the PolarFire 2 later this year. It's software ecosystem is already out. And we got customers waiting for those parts. In fact, all the initial runs that we're doing even for samples are already spoken for. So there's a strong demand for that part, even outside the A&D sector as well as in the A&D sector. So I think we're already doing that. We're only selling lots and lots of those devices in the markets outside of A&D.

Your next question comes from the line of Tore Svanberg with Stifel.

Congrats on the results. Steve, you talked about lead times starting to extend. I was hoping maybe you could give us a bit more color there. Is it more broad-based? Is it specific products? And just based on the current environment, do you expect lead times to continue to extend as we go into the second half of the year?

So I think, in general, lead times are broadly expanding. And I think in another quarter or so, there could be nothing available in 4 to 6 weeks. I think it's entirely possible. Today, we still have -- when we say we have 185 days of inventory, even though a lot of that inventory, we keep it in the die form. But if the demand comes up in the exact mix where we have the die, we can still assemble, test and ship it in 4, 5 weeks. But that's not going to last very long. I think over the next 2 quarters, definitely, the inventory goes down significantly into the range of our long-term model. And I think many, many products, die inventory would really be fairly low. So the lead times could see a broad-based expansion in the coming quarters.

Very good. And as my follow-up, you talked about the 2 pillars moving to 5 pillars. I appreciate you're still doing all the work there. But I just want to make sure I understand some of the compositions obviously, networking is straightforward. But you mentioned HPC and Edge AI. Edge AI is pretty clear. But HPC, would that primarily be then storage for infrastructure? Or is there anything else that goes into that pillar?

The high-performance compute has really microprocessor business, our FPGA business, our timing business, those are most of the products in there.

Your next question comes from the line of Quinn Bolton with Needham & Company.

I guess I wanted to ask, you mentioned a couple of times in the script that you're starting to see the [ distis ] come back with larger orders. I guess, in the guidance for the June quarter, are you assuming some level of restocking? Or do you think that restocking happens later in the calendar year? And then I've got a follow-up.

So we are expecting some level of restocking, but you should think about that, but that doesn't necessarily mean that the months of inventory or days of inventory in distribution will go up. So we're guiding 11% up in revenue. And if you take a distributor and say they've got $50 million of Microchip product, and they are growing at that rate, we'll call it, 10%. They have to grow their inventory by $5 million just to keep their MOI flat if their sell-through is increasing at that rate.

So yes, we are expecting some restocking. The level of that, I think it's hard to predict, but I think it's needed. Distribution inventory is really low at this point in time. And in a growth environment, they got to position theirselves to be able to support their customers.

Got it. And then a quick follow-up for Steve. It sounds like, Steve, you feel pretty comfortable you can get the external foundry waivers. But any tightness on either OSAT, substrates, especially as you think about the PCI and the memory and storage controller business. Is there any tightness there?

Yes, of course. I mean like I said, there is tightness everywhere. Our foundries, and all major foundries we do business with, probably 70%, 80% of the process technology nodes are constrained. When I say constrained, I would say they are very tight. If I need a couple of hundred more wafers, can I work the system and be able to get it? The answer is, yes. But it isn't like you're sitting at a 80% utilization. And if I wanted more wafers, I have them right away.

You kind of got to work the system and make the calls high up and escalate it and request it and somebody else in the system has a downside. Then you get it, you may not get it today, but may get it 3 weeks from now. So everything is basically full. And I think that's the system we're looking at it. But we have a higher allocation as the future quarters come in because foundries are adding capacity. So we have increased allocation from them. And with that, the business will continue to grow. But like I said earlier, can I grow a business 50% in 1 quarter? No. But can I grow at the rate we're talking about? Yes, and there's even room for upside.

Now when you come to the most advanced capacity like 3-nanometer node, that's totally different equation. There the capacity is less than half of what the world requires.

In substrates?

In wafers.

Sorry, just advanced packaging substrates, any impact on the -- like the data center business?

Yes. So we are constrained on substrates. We're qualifying additional substrate suppliers. We are working with the existing suppliers to get more allocation. It's just like the wafer, substrates are very tight. One problem with the substrates is they have a shelf life, right?

12 months.

Yes. So they have a shelf life, and therefore, you can't build a whole bunch in the down cycle, thinking you can use them 1 year, 1.5 years from now. They have a shelf life. So you have to really buy it in time. Are substrate constrained? Yes. Is it impacting our data center business? Yes. Is it impacting our connectivity and networking and automotive business? Yes. And we have delinquency, our nonsupported dollar volume has gone up significantly, but it's not at a crisis level. It's not like it was post-COVID. We're working through it. And there are no real screaming customers so far. I think we're working through them.

Your next question comes from the line of Joe Moore with Morgan Stanley.

I wonder, you talked about your discussions with customers and kind of rebuilding relationships that had been strained during COVID. Can you give us some color on what those conversations are like and I guess some of the strain was enforcement of LTAs, price increases, things like that. How does that inform? How you're going to deal with those sort of same considerations in the next upturn?

No, I think our conversations in the last year have been very good and we visited a lot of customers, a lot of customers visited us, and we met them at events around the world from CES, to MASTERS conference, to other industry conferences. And we owned up to some of the mistakes we made in the past. And customers, engineers always liked us. We had a very good product. We're a high-quality product. We were reliable suppliers, but we just used some policies that turned off purchasing managers and manufacturing people and others. And we made commitments to really work with them better. And I think that has gone well. People have given us a chance.

And now what's happening is we are seeing our competitors hurt them with large price increases and all that. So they're even coming more to us. And since we haven't increased the prices, we're gaining share. So I think all that is kind of working well. Our competitors are actually helping us right now.

Your next question comes from the line of [ Janet Ramkissoon ] with [ Quadra Capital ].

Steve, thanks for the update on the nine-point plan. Just my question about substrates has already been asked. But just one on the business unit realignment. If you look at the 3 markets that you've added to the basic business, microcontrollers and analog both businesses have pretty large TAMs. And given the -- what's going on with AI and data center growth those markets could present significantly larger opportunities. At what point do you think that you might have to revisit your margin targets? Or am I being too optimistic about the growth prospects of these 3 new areas?

When you say -- Janet, first of all, thanks for asking the question. When you say we have to revisit our margin target, are you concerned our margin targets are too low or too high?

Too low.

Well, that will be a high-class problem. Let's -- when I came back 1.5 years ago, our gross margin was 52%. And when we gave a target of 65%, I've been fending the blows for 1.5 years, like why our target, how will we get there? And we were so low and this and that and competitors are doing this and competitors are doing that. Now it is in the striking range, if you take our underutilization and add it back, I think we get there. So let's get there and then if you're...

Yes, the reason -- I just want to add this. We look at a company like Anthropic, they went from $1 billion in revenues, and now they're talking about potentially exiting the year $45 billion. I mean this whole AI business has just taken off like a rocket. And we've never seen growth in any market like this ever. And so if you look at AI on the edge. No, Anthropic, they're one of the large language model providers. And you look at the growth there, you look at the growth in OpenAI, it just seems to me that if you -- that you might need to be preparing to keep up with a higher level of growth rate than you have been used to historically. That is just being a long-term shareholder, I just thought I'll throw that out there.

We'll welcome that opportunity. Yes. I mean we are working on it. We're developing the products. We see the market. We see the opportunity. That's what you've heard from Brian now in 2 out of the 3 last meetings and last meeting, we talked about T1S. So we are engaged. But I think it's premature for us to sign up to anything different than what we're signing up for.

Yes. Okay. I wasn't asking you to sign up. I was just trying to think just sort of think outside-the-box to try to figure out where your head might be.

Your last question comes from the line of Blayne Curtis with Jefferies.

I actually wanted to go back to a couple of the prior data center comments. I think you broke out at one point that within data center compute, data center was about 80%. I wanted to make sure that's the right way to think about it for the fiscal year. And then if you look at a lot of the peers with exposure -- analog peers with exposure to data center, that business is growing substantially double digits for the last couple of quarters. Just kind of curious, I know you probably don't want to give the exact numbers, but is that the kind of growth you're seeing? And is there any headwinds still in that segment that are going down.

So I'll start by saying I don't think we've given a percentage of our data center and compute what data center makes up. But it is the larger piece for sure. And then within that, there's various different areas that make that up. Brian's business unit Data Center Solutions is one of those, which we have high expectations for. I'll let Brian speak to the rest.

Yes. We're also seeing high growth in our digital power control for many of the power supplies that exist within data centers within our power group.

If you look at our Data Center business unit, Brian's business unit, year-over-year has a huge growth but we're not quite willing to break it out unless we do that as part of the overall breakout that we have talked to you about and we're analyzing it. The problem is the data center one is not as much as of a challenge. Some of the others are a significant challenge, like connectivity and networking and timing business and others because some business units have some parts which are analog, some have a microcontroller core, so the portion goes into microcontroller, is something IoT or is something networking? What is the difference? They look very close. Is it a data center business unit? Or is it a part that goes into data center, but it's a power management part that comes from AI -- I'm sorry, comes from analog business unit.

So we have content in data center market or data center servers, which are microcontrollers, which are power management parts, which are...

Ethernet, USB.

Ethernet, USB, so there are various business units, probably 7 or 8 business units contributing to our content inside the data center. Brian is just one part of it. With the storage controllers, PCI Express and the third one was memory, CXL and those. So it's a complicated breakout. And we are somewhat concerned about confusing you and not really helping the cause. So we're trying to figure that out and put it all together so we can share with you.

And then just a quick one for Eric. I know you have a big piece of variable in your OpEx. I'm just kind of curious how to think about OpEx for the fiscal year.

Yes. So you're right on that. So when the business is doing well, we have quarterly bonus programs for our employees, which were very low in the down cycle and now that we're coming back, those programs are coming back nicely and guiding 11% up. We're expecting it to be another good quarter for our employees from a bonus perspective. And so that's why you're seeing that along with other compensation elements that were kind of pent-up that we're bringing back to employees drive our OpEx dollars up in the current quarter. I think at the midpoint of guidance, it's about a $15 million growth quarter-on-quarter in OpEx.

And the nice thing about these plans is they are quarterly and we can modulate them in good times and bad times. High-growth quarters, lower growth quarters to keep operating expenses in check. So showing good progress there, but are making investments that we think are important to reward and incentivize our employees to stay with us and continue to produce great work.

This now concludes our question-and-answer session. I would like to turn the floor back over to Steve Sanghi for closing comments.

Well, thank you all for joining us today and hanging in there with us. We'll be attending several conferences coming up this quarter, I think, starting in early June. So we'll see you on that circuit.

Yes. We will actually be the first one. I think it was a JPMorgan conference in 2 weeks and then a number after that. So we look forward to seeing everybody.

Thank you.

Ladies and gentlemen, thank you for your participation. This does conclude today's teleconference. You may disconnect your lines, and have a wonderful day.

Good afternoon, everyone. I'm Matthew Prisco, analyst at Cantor Fitzgerald, covering semis and semi cap. Today I have the pleasure of hosting Eric Bjornholt, CFO at Microchip; as well as Sajid Daudi, Head of IR. So thank you both very much for joining us today.

Thanks for having us.

I wanted to open this chat with some near-term level setting. Last quarter, Microchip highlighted a number of positive developments in terms of cyclical indicators, one being booking strength as December quarter moved significantly higher and strength into January. Can you offer some color around any particular areas of strength within those bookings? Any particular areas that could be lagging? And then how have these dynamics develop through February and early March?

Sure. Sure. So I'll start by doing the safe harbor and saying I refer you to our filings with the SEC that identify important risk factors about the company. And we'll be making some forward-looking statements during this presentation and discussion.

So bookings have been really strong. So both January and February bookings were very, very strong. It's the strongest first 2 months of the quarter that we've seen since like the first quarter of fiscal 2023. So bookings have continued to be good, and that's even with the Chinese New Year falling in there. So it's been really good. Backlog is building nicely, obviously, filling up the current quarter and then building our visibility into the June quarter, giving us confidence.

The bookings have really been broad-based in nature. It's not one particular end market or geography that's driving it. It's broad-based. We think there's a lot of customers that are now coming back. They've depleted inventory or now having to buy in line with what end consumption is. So that's good to see. We have some secular drivers that I know we'll talk through as part of this discussion. We are seeing strength in data center aerospace and defense and some of our networking areas are also quite strong right now.

Okay. And I know you guys talked about expedites increasing significantly and lead times extending in spots as well. So -- can you offer us some more detail on specific areas where these dynamics are most prevalent today? And maybe how you expect these trends to develop over the coming quarters?

So what we're seeing in expedites is customers are placing orders and they might place them out in April and then try to pull those in, into March because once they know that we've got them started in our manufacturing line, we have capabilities to have pretty quick turns to them. So we're seeing quite a bit of that activity that's continued last quarter into this quarter. Second piece of your question was?

And then lead times, lead times are extending. And how do you think these dynamics kind of progress from here based on just what you're seeing in customer conversations?

Yes. So lead times are generally quite short, but there are some pockets where lead times have extended out. Everybody knows that 3-nanometer as an example, some of the more advanced process technologies are pretty full at the foundries. So we're seeing extension of lead times there. Also some of the substrates impacting some of the assembly and test portions of our business where we subcontract that out to a third party have extended, and we're reflecting those in our lead times, being sure that we're communicating with customers. They know what they are. But I'd say for 90% of our products lead times are still very short. We've got 200 days of inventory on the balance sheet and more capacity than we need at the current time. So I expect lead times for a lot of the products to stay quite short, but there's going to be these pockets and it tends to be that, that it will expand a little bit over time. But right now, it's a pretty small subset of products.

Okay. And maybe on the inventory front, how should we think about the duration of the customer inventory, I believe, in your direct customer inventory? I know from a [ disti ] side, it sounds like things are pretty good there. So how do you think about the customer side? When does that start really tracking your shipments in line with demand?

Yes. So as you said, distribution, we think, is pretty much corrected at this point in time. Last quarter, we ended with 28 days of inventory in the distribution channel, which is normal. The difference between what was sold out of distribution and what was sold in was only about $12 million last quarter that sell-through exceeded sell-in. And those might even normalize to be nothing. This quarter, we'll see how it plays out.

On the direct side, I think there is still pockets of inventory that are out there. We don't have great visibility into that other than through customer discussions and discussions with our distributors in terms of what they're seeing from their end customers, but we are watching things in distribution in terms of distribution sell-through. We would expect as that inventory depletes from their customers, the distribution sell-through will have to pick up. And in turn, what we're selling into distributors will rise with that.

So we look at sell-through very closely. And what we are seeing in many cases that we have new bookings activity coming in from customers that haven't bought from us for some time, and that's a clear sign that they've depleted out their inventory and need to come back and start to buy in line with end consumption.

Many customers buy multiple products from us. And so it doesn't happen for each product at the same time, but this gradual increase that we're seeing, I expect that to continue for the next few quarters.

Okay. Let me move into the segment side, starting with the industrial market, which accounts for roughly half of your revenues overall. How have you seen the recovery play out here today across the subsegments? And what gets you most excited in terms of growth here over the next few years?

So industrial is a big piece of our business, including aerospace and defense, which is about 18% of the business. We include that in Industrial. In total, it's about 48%. Aerospace and defense has been a really strong segment for us for quite some time. We are a large supplier to the Department of Defense. And obviously, conflict around the world isn't great, but that is a benefit to our business. And so we are seeing some upsides there and it's been very steady, high-margin business for us for quite some time. Other areas of industrial, and Sajid can chime in here also, but we're seeing anything, Industry 4.0, factory automization, robotics, things like that. Medical even is doing quite well. Anything else you want to add?

Yes, some of the edge IoT type of activity and FPGA plays on that. Industrial content growth, generally speaking, is increasing as well. And so other products are finding space there, too, in the value chain.

Perfect. Maybe digging a little bit deeper into aerospace and defense, given the current geopolitical landscape, can you give us any help in breaking down the revenue pay between those 2 markets? And -- where is Microchip best positioned? And have you seen any change in order patterns or customer conversations over the past couple of weeks?

So A&D has 3 broad categories for those defense, there's commercial aviation and there is space. And defense is the biggest piece of it. We don't break it out specifically. But obviously, the defense side is where we're seeing really good traction right now. Space also, there's a lot of things that are being launched up into the atmosphere. You've got new space, which Sajid again could talk about a little bit more if you'd like him to dig into it a bit. But we're seeing good activity there. Commercial aviation is also strong. We sell a lot of products into all the planes that are going up. And with that, our content has continued to grow. FPGA has a strong position in aerospace and defense for us, and that's been a really strong area of growth for us.

Okay. Perfect. I believe you've recently highlighted these new product introductions, which should start ramping in 2026 and bolster growth for Microchip. So can you maybe walk us through this dynamic? Any particular area that these wins have been accruing more than others? And how should we be thinking about the magnitude of these ramps and how that can contribute to outsized growth versus the market in '26? And does that continue through into '27?

Yes. So -- yes, absolutely, right. So we have actually two very significant platform level products that are ramping into '26-'27 time frame. And these are general and architectural changes rather than kind of a one-off product technology change. And so the first area is the Ethernet side, which is the automotive Ethernet side, and this is our 10BASE-T1S product. And here, we already have actually a platform design win with Hyundai Motors, which we announced last earnings call and additionally, some other kind of Tier 1 OEMs. Basically, these companies or these customers as they're looking to modernize their legacy architecture, which has traditionally been LIN and CAN [ archit ] nodes into the Ethernet side. That's really where this product kind of comes into play. And we feel that as we kind of progress to a zonal-based architecture into the next-generation EVs and a ADAS platform, the 10BASE-T1S product category, that's a pretty prevalent lift in here. It's a refresh cycle. It's a once in a multi-decade type of transition that's happening here.

And so all these legacy nodes will be eventually replaced by 10BASE-T1S on communicating in different areas of the car. So very exciting. And again, early stage. We just are ramping in there. So I think revenue contribution will be minimal in the '26 calendar time frame, but really ramp -- start ramping in the early, mid-, late '27 and beyond standpoint. And this is like I said, a multiyear trend that we think is happening.

And this isn't replacing anything you currently do. This is all fresh...

This is going to be all fresh, right? So this would be existing. So we're not going to cannibalize. Some of the existing legacy nodes. Obviously, we have a presence there, but then this will be incremental there. And then the other thing exciting thing that we have on the data center side is our -- so we now have the first production-ready Gen 6 PCIe switch on 3-nanometer. And this is something that's been -- again, Microchip has a long history of acquisitions and through our last one, which is Microsemi, we acquired this piece of the equation, and Microsemi was rolling up a whole bunch of players as well. So this kind of legacy goes back to the PMC-Sierra days for those who are familiar.

And so Microchip today is really the only company that has -- so PCI is not a new area for us. So today, we're serving it with Gen 3, Gen 4, Gen 5 and Gen 6. Gen 5, we had a little bit of misstep on where we tried to develop some SerDes in-house, and that delayed that product category. So we lost some share there because of that delay. And so for the Gen 6 launch, we went a technology improvement there from where competitors are on 5-nanometer. So we went 1 node lower to 3 nanometers. And what that does is just there creates quite a bit of value proposition where our product just going from a technology difference drives 15% to 20% greater power efficiency or better power efficiency using our products.

And again, as you kind of think of a gigawatt style data center, that's substantial savings and when the GPU clusters are expanding to no end. So that's another area. And again, like I said, that's ramping. We announced this. The product went to production a couple of quarters ago. And so last quarter, we announced 3 design wins on it as well. And so again, that's going to ramp minimal in the '26 time frame, but more so in the '27 area and again, a multiyear trend there. And our opportunity there is kind of multifold, right? A, just again being a technology leader, so just a pure play a share gainer from some of the established players in there. And then the other part would be as a dual source partner for a lot of these customers because as they look for optionality, that will be an opportunity as well.

And so -- like I said, today, we're the only one that's providing that entire suite of PCIe switches, which becomes a critical piece. One point I'll make is, even as you think of some of these new generation areas like robotics and stuff, a Gen 4 PCIe switch is more than capable to deliver some of that. So it doesn't need to be kind of the latest, greatest. And typically, what you see is hyperscalers as the first adopters of the latest generation. And then quickly followed by 18 months or so with enterprise adoption and stuff. So we think we have a good runway ahead of us in this area. And certainly, we have to kind of reestablish ourselves a little bit, but certainly, those are the 2 areas that we're pretty excited about, and the ramp is going to start coming through in the '27 area, more so significantly.

And on this PCIe, this is roughly a $2 billion market today and industry estimates are it's going to grow to about $10 billion by 2030. So the growth opportunity is large here.

Perfect. And I definitely want to dig more into data center, but just to wrap up Industrial quick. You guys made an interesting comment recently saying that hardware is becoming a little bit less important versus the tools and software being offered to customers and enabling the intelligent edge. So firstly, does that mean that the hardware such as the MCUs becoming generally kind of commoditized at this point with a broad array of offerings that are good enough? And if not, how does Microchip differentiate there? And then second, where does Microchip stand with the toolings and software? What are you offering to customers here? How should we think about investment focus in the space from the company? And maybe how do you differentiate from the competition?

Yes. So the hardware is still a very important element. You've got to have the right product that does what the customer needs you to do. But software is a growing element and we are doing lots of things to make the customers' journey and use of our products easier and provide more tools to them. We're doing all sorts of things internally with AI and machine learning, making some of that available to customers where we've trained this stuff for 5 or 6 years internally and now opening it up and they can use this. And from a tools perspective, from a design perspective, to save 30%, 40%, 50% of design time by using it. And that just -- everybody is limited on resources that they have on the engineering side. And so anything that we can do to help them with that is extremely important. Sajid, would you add anything else?

No, I think you kind of captured it well.

I mean is that bolstering the ASP side? Is that bolstering the margin side? Is it part of TSS? Like how does this kind of change Microchip story, not just today, but kind of looking out over the next 3, 5 years?

It's all those things, right? So TSS is selling total system solutions, and we've been working on that through the acquisition strategy that we essentially executed over about a decade. And so you go in with a base kind of anchor product, and that could be an MCU, it could be an FPGA, it could be MPU. And that's kind of the brains of the system. And then there's all this other stuff that goes around it, whether it's analog, timing products, security products, memory products, all these things. And we offer all that to our customers. We provide them working reference designs for that and then speed their time to market where they're not trying to get chips from 6 different companies to interact well with each other. We do that for them and give them a head start and bringing the market quicker. So that's a strategy we've been employing for the last decade or so, and it's just getting more refined every year, and we're helping customers save money.

Okay. Now shifting over to data center. You already touched on the PCIe Gen 6 opportunity there. I think the last quarter, you had said 3 customers. And I think intra-quarter, you guys said 4 customers. How do we think about kind of that pipeline converting from here? Can you, I guess, first give any color on those 4 customers that you have? And then how do these new customers roll in? Is it kind of proof of concept with these early customers to make sure it's working. And then we'll see in 6 months here is the next tranche of customers? Or -- just how does that kind of work?

So we started sampling back in October time frame. And it takes time. These are complex products. We have many designs that are in various phases right now. We've talked about 3 or 4 wins that we've had there. One of those wins is material from a revenue perspective that it's expected to be $100 million plus in calendar year '27. So that's important. And we've got, again, 20 or so others that are in various phases of design and we've got a product that we think has a very significant competitive advantage over the rest of the competition from a power consumption standpoint, what Sajid mentioned before. In these gigawatt data centers like 8% of the power is consumed by PCIe. So if we can provide a 20% advantage there a reduction, it's a meaningful advantage and people are strongly considering our products. So we're super excited about it, and we think the growth opportunity is quite large.

Yes. And even -- I'll just add to that is that even on the PCIe side, you almost vendor-agnostic, right? So you can have a media style GPU or an AMD or custom ASIC and still be able to use our PCI switches. So certainly, we're today sampling across the board with everybody and some. So really very healthy kind of sampling activity going in there and timing will just be defined by as these customers start coming in. But today, the reach is very strong, and the response has been really good. So we talked about 15%, 20%. I think if you really talk to our business leader there, you'll say field tests are coming back even more positive. And so this is what everybody is trying to solve for is what the power efficiency in the data center. And so we certainly have a solution there, and interest levels are very high at this point.

And we have more products coming over the next couple of months, we expect for our retimer products to come to market also that can sell alongside the switches.

I guess maybe how should we be thinking about Microchip's ability to maintain its advantage to transition to the Gen 7 PCIe switch. I think that's theoretically coming out this year. I mean, what type of improvements should we expect gen-over-gen, you talked about the 15% to 20% you're seeing now? Is that something similar next time around? And then maybe what are the technology shifts that would drive the customers to kind of adopt this next-gen offering?

So I think we're in a very good position because we're already on 3-nanometer, right? And Gen 7 will also be on 3-nanometer. So since we were the first there, I think we were a step ahead of the competition, which will likely transition to 3-nanometer and next gen, and it's always most difficult that first time that you're bringing up a new technology. So that's an advantage, and we absolutely expect that product to be out in the next 12 months. Anything you want to add to Gen 7?

And I think just the fact that we are today kind of have all of the generations in there. Once you get designed in the first 18 months or so is really where a lot of this design capture is happening. So once you get designed in and you can deliver to the customer that you have a viable product, and the next generation, the incumbent always just has an advantage, right? So it adds to that stickiness aspect to that relationship. So I think -- and again, I will just emphasize like today, we're really the only ones that are offering that Gen 3 to Gen 6 type of activity. And the product road map goes well beyond Gen 7, too. I think you talked to the team there, they have it scoped out for a very long time. And I think by Gen 10, Gen 11 is when you get to just a direct Ethernet connection at that point, right? So there's no intermediary in that area. But certainly, yes, long tail ahead of us along...

How about looking at the retimer opportunity to pair with these switches. Can you talk about expected adoption trends there? And what would drive the customer to choose the Microchip part here and maybe help size this opportunity versus whether it's switch, whether data center revenues?

Yes. So it's really complementary to the PCIe switch, right? As you kind of think of what a switch does, it's kind of giving you that communication between a GPU to GPU or memory controllers, this really kind of extends that communication out further distance, right? So it's really a complementary product there. And again, I think as you think about the opportunity size here, like Eric was saying, the retimer business is an incremental $2 billion to the $10 billion TAM that we said the industry size in there. So again, a smaller area, but goes pretty hand-in-hand. So that -- think about that total solution systems that we're providing. It just helps address a bunch of different areas for a customer as they're looking to design new products. So I think, again, the product is expected to be out next month. I think there's obviously active dialogues with existing customers that may be looking out to kind of trial it out and then go from there. But I think once we have more data points around what the customer feedback is and once it's out in the marketplace, we'll have more color to share. But at this point, we think it's very complementary and expected to come out very soon here.

And just to clarify that, the $2 billion that Sajid mentioned, that's a 2030 number that you're talking about for TAM, not today.

Okay. So $10 billion from the switches, $2 billion from retimers?

Correct.

Okay. How outside of the switch and retimer dynamic, what are other areas in Microchip's portfolio that excite you on the data center side of the business? And I don't know if you could help us understanding the composition of that portfolio.

Yes. So I can start, Eric. So it's really kind of -- it's a full stack solution. So obviously, we've got the switching connectivity side as kind of the one stack there. Then on the storage side, we have kind of RAID-on chip controllers. We have the NVMe, SSD storage controllers. We have the IO switches there as well, then included kind of beyond that as well as the power management products that we offer. So voltage regulators and again, just kind of balancing what the density of a lot of these required density requirements for a lot of these data centers are. So we have a lot of tools on the power side. Then kind of extending it further, you get into the clock and timing side of the equation, when you look at Root of Trust activity and all that, we have a lot of product offerings there as well. And so really a full stack solution and a kind of depth of product areas that we touch the data center with. And again, it's not limited to just the hyperscalers. Obviously, the enterprise data center side, which is a refresh that everybody is kind of waiting to happen because all the spend is going into the AI side today. But I think those would be the areas that would be kind of areas to highlight for the data center offerings.

When you think about those other areas, is there a share gain story we should be thinking about here? What's your market positioning? How should we think about growth? Is this just going to be linked to hyperscale CapEx? Is it going to be linked to something else?

It's probably the best way to think about it it's more linked to the size of the market, but we have competitive solutions in all these areas that Sajid talked about. And when you win some of these larger sockets, it opens up doors for you to have the discussions with these customers about where else we can win.

Yes. And that, I think kind of gets us to an interesting dynamic as well. So anybody that's been following Microchip, you always hear kind of 2-pillar story, right, where there are Fortes like microcontrollers and analog products primarily. And now today with connectivity and some of these other areas, these are kind of new pillars that are coming and getting established, meaning this is where customer dialogue is starting at and then you have a lot of attach opportunity based on what the customer wants to design. So Microchip, you're starting to see a shift from the kind of 2 key pillars to 5 pillars, which is connectivity, data center and AI/ML on the software tool side. So certainly just an expanding growth story here.

Perfect. So gross margins, they came in nicely ahead of expectations last quarter with decline in inventory reserve charges and better product mix. And maybe starting on that inventory reserve charge side of things. How should we be thinking about that number kind of normalizing into the March quarter? And what gives you confidence for a rather meaningful sequential decline there? And then how do we think about that kind of level throughout the remainder of the year?

Sure. So maybe just to highlight, previous quarter, the December quarter had kind of an abnormal event in it where we had an IP sale within our licensing group that added another $20 million, $25 million of revenue at 100% gross margin. So that helped the upside in gross margin that we saw compared to maybe what The Street was expecting, although it was in our forecast for the quarter. This quarter, we're still showing gross margin improvement. I think it's going at the midpoint of guidance to 61% compared to 60.5% last quarter. And that is with that licensing business not repeating in the current quarter. So we had a bit of a headwind on that front and still showing nice growth in revenue and margin.

Specifically on the inventory reserves, they were -- so new inventory reserves were about $58 million last quarter. I kind of like to view it that when things aren't normal, inventory is rightsized, demand is normal, that it's probably $15 million to $25 million a quarter. So we've got a continued benefit that's coming as those reserves reduce. I think we're pretty close to normalize this quarter when you combine what the new reserves are going to be plus what is likely to sell through of some inventory that was previously reserved based on our accounting policies. So with the 61% margin that we're forecasting this quarter at the midpoint, I think inventory reserves are pretty much normalized in that number.

Now on top of that, we've got underutilization charges in our factories of about $51 million, $52 million per quarter right now. And those are going to take longer to come down because we need to grow back into our capacity. And I'm projecting that's probably going to take us a couple of years to burn through those, but they should be coming down modestly each quarter as we move through the next 2 years.

Is there anything we should look to, to inform us of that decline? Like I know it has to be the internal products that grow for obviously utilization to tick higher. So what focus -- what areas should we be focusing on to kind of tell that your utilization rate is going to climb higher than your utilization charge can come down?

Yes. So I mean, we will be pretty specific with you on our earnings call and what our expectations are, but we need to ramp the factories. And we only do about 35% to 40% of our production in-house from a wafer fab perspective. We outsource the other 60% to the professional foundries. So we need growth on the products internally. Our internally produced products, which are kind of our standard microcontroller analog products, some of our timing products, some of our memory products get produced inside, so we need those to grow. Now a lot of the secular things that Sajid was emphasizing here as growth drivers for us, whether it's the data center products, our FPGA products, some of our Ethernet networking type products, most of those are outsourced. So we need growth on kind of the core MCU analog products to grow back into our capacity, which will happen over time, but it might not be as fast as growing as what we see coming from some of these secular areas where we rely on the foundry. So it takes some time, but we're confident we'll grow back into the capacity and have really high confidence about our 65% gross margin target that we have on the non-GAAP side. We've been giving some data each quarter about, hey, if we didn't have these inventory reserve charges under utilization, the underlying product gross margins are like 67%, 68%. So the health of the product line is there.

The good thing with some of these growth areas we've talked about in FPGA, in data center, in Ethernet, these have higher than corporate average gross margins. So they aren't going to help us fill the factories as they grow. But we'll have a bit of a tailwind on the gross margin side because they're highly complex, high-value products to our customers.

Perfect. How about the pricing side? How are you thinking about Microchip's pricing power today, particularly we've seen a number of analog pricing increases year-to-date. So how do you think about that today? And maybe moving forward, as we think about that annual pricing decline of low single digits typical in the industry. Any changes in the customer collaboration you have, the TSS, we're talking about all these dynamics they can kind of change that?

Yes. I would describe our pricing as stable. You're right that many of our competitors have raised prices. We were actually one of the last of our competitors to raise prices in the last up cycle. And you know one of the things when Steve Sanghi came back as CEO was he really wanted to focus on customer relationships. And so we want to be very careful with what we do from a pricing perspective and not gouge customers in any way. So there are clearly some increases that we are seeing in cost in the supply chain, right? It could be gold costs are up or some of the assembly and test subcontracting costs are up. And we will likely do something that is more specific with customers and be very transparent that, "Hey, if we have to raise your ASP by 3%, this is why, and here's the details behind this."

But we're not really thinking about a broad-based increase in pricing of the customers. We want to -- our gross margins have improved dramatically from the low that they were at last March at 52%. We're guiding to 61% non-GAAP gross margins. So margins are moving upward very nicely and getting closer to our target. We're not going to use pricing as a lever to get there. We want to focus on winning more designs at customers and gaining -- having them gain their confidence in us that Microchip is going to treat them the right way.

Got you. Okay. So with a minute left, I got one last question here, I'll give it a quick one. You talked last quarter about kind of increasing transparency in end markets and maybe something new is coming up. I guess, first, what's the impetus for doing that today? And second, what should we expect to see here in May when you kind of come out with the new thinking?

So a lot of our competitors give more details on a quarterly basis than we do on end markets or certain growth areas. And we've kind of used a once-a-year process to do that. So we've been refining our systems and processes and finding a way and discussing it as an executive team, how best to update investors in a consistent manner. So I can't share the details now. We'll share that in our May earnings call for the March quarter, but it is likely that you're going to see more information to allow you to gain some insights on how Microchip is growing in some of these secular areas because they get bundled into 3 categories. It's microcontroller, analog and other. We know we need to give more information on data center on FPGA as an example, on networking. And we'll do that. So we'll define that more and more to come in early May.

Perfect. Well, thank you guys so much. Really appreciate it.

Thanks for having us.

Thank you.

Welcome back, everybody. I'm Joe Moore, Morgan Stanley Semiconductor team. Happy to have the management team of Microchip, Eric Bjornholt, CFO; and Richard Simoncic, COO. So thanks, guys, for coming. Really appreciate it.

Thanks for having us.

A lot of interest in Microchip these days. You guys have probably the best rate of change off the low point and a lot of faith in Steve's ability to kind of get the business reoriented. You guided March up 6% sequentially, quite a bit stronger than seasonal.

Can you help us understand that? Is how much of that is customers not depleting inventory anymore? Any indications of restocking, half of the customers destocking and still like just where are we in that dynamic?

Sure. So let me start by saying during this discussion, we'll be making some forward-looking statements about the future financial performance of Microchip, and we refer to our filings with the SEC that identify important risk factors about the company.

So yes, we're guiding the current quarter up 6.2% at the midpoint. I think it's a combination of things. We've got good product momentum in various areas, which Rich can talk about as we go through some of your other questions. But bottom line is distribution inventory is kind of normalized at this point in time.

If you look back a year ago, we had like a $100 million difference between what we were selling to the distributors and what they were selling through. And now that's collapsed down to only be about $12 million last quarter.

I think customer inventory is still normalizing a little bit at the end customer and at distributors' customers. So there's still a little bit to go there, but the distribution is completed at this point. And then we've got, as I said, some pretty good product momentum in certain areas that we'll touch on throughout this discussion.

Great. And you've talked a little bit about visibility building. You talked about a good January bookings month when you reported. You talked about June backlog looking better, getting that forward-looking visibility a little bit. How much of that feels like it's end demand related? And did that persist after the Chinese New Year holidays?

So I think the demand environment is definitely improving. Our February bookings were also quite strong. And if you look at the first 2 months of the current quarter, they are the highest first 2 months we've seen in bookings since, I think, the June quarter of fiscal year 2023. So booking...

There's a lot of math when you say stuff like that, I got to go back.

All right. I got to make sense of it, it's been a long time.

It's been pretty good, yes.

And it's been good. So backlog is filling up nicely, giving us confidence, obviously, in the current quarter guidance and then giving us better visibility into June and beyond.

And are customers wanting you to have that visibility? Or are they still believing that you can turn stuff in a pretty quick fashion?

They mostly believe that we can still turn it in a pretty quick fashion. We've got 200 days of inventory still on the balance sheet. So we -- and our lead times are generally short.

There are some anomalies to that or some instances where lead times have extended, if it's on very advanced process technology, it could be some things with substrates and some of our back-end assembly and test manufacturing. But generally, lead times are still short.

Okay. And then you mentioned distribution inventory looks cleaned up, but direct is mixed. I guess that seems a little bit more cautious than what we're hearing elsewhere. Some of that due to maybe the excess inventory build that you guys saw at the last peak. But just what's your visibility into that starting to get cleared up?

So you know we have this program called PSP, the Preferred Supply Program. And with that, many of our small to midsized customers didn't get inventory when they wanted it. They had to wait because that capacity and inventory was consumed by those that came first for the product.

And so with that, when they finally could get product, they bought more than what they needed, right? They did not want to get caught in that situation again. In some cases, bought a year's worth or 2 years' worth of product, and that's what's taken this long time to really drain through.

But I think we see significant progress in that. We're waiting to see a nice inflection in sell-through from distribution. That's the next thing we're really looking for because now distribution inventory is corrected for the most part. But when there -- are their customers now correcting or in the process of correcting, that will lead to higher sell-through and the distributors will need to buy more.

Helpful. So when Steve was here last year at this time, he walked us through a 9-point strategic plan, and we're kind enough to orient that around our conference, which we appreciate. The numbers is part of it, we know, the inventory aspects and gross margin and things.

But how are we progressing on the customer relationship improvements, the sort of product focus and those elements of the plan at this point?

I'll let Rich speak to customers and products.

So on customer progress, we really revamped. We changed our whole ethos and how we actually work with customers and talk to customers. We really sped up the rate at which we are releasing products within the company and getting answers to customers.

And Microchip for a period of time there, had become arrogant -- right? I don't know how to say that. And we had to fix that whole relationship with customers. And so that has significantly changed over time.

And then at the same time, we're doing that, working with customers on where do we need to move to. So we really doubled down on wired connectivity, ASA, camera connectivity, robotics. We launched PCIe Gen 3 and Gen 4, believe it or not, 12- and 24-lane switches for industrial applications. And all of those are taking off and doing much better in the marketplace, providing much better growth than a more traditional microcontrollers and basic analog products.

You have a very rich business portfolio, all the acquisitions that you did we know all the pieces that are in there. I guess if you went through that period where you were maybe a little bit arrogant, are those cultures still there, the richness of the product portfolio is still evident. We'll get into some of the specific drivers, but how do you feel overall about the Microchip portfolio? And are you going to give us at some point more -- we sort of have the 3 segments, but like...

So there was quite a bit to fix over the last 2 years. We reorganized the company to remove the silos of technology, and we've built it into now just 5 pillars, right? So it used to be 8-bit and 32-bit and 16-bit and everything was separate, right?

And so now we've changed the organization. So all microcontrollers are run under one person, right? And all of compute is under one person. All of connectivity is under one person. All of analog is under one -- and then we have those groups then working together based on megatrends to put together reference designs and bundled products that actually work together.

And so we got rid of some of the personalities that were in there that were preventing us from actually working as one cohesive team. And that's actually really moved our product development substantially.

Okay. That's helpful.

So I think you had a second part of your question there that was kind of talking about how we disclose things today. Right now, we disclose microcontroller, analog and other as our product lines, and we are working internally to find ways to better and more frequently update The Street in terms of how some of these faster-growing areas are growing because right now, it all kind of gets lumped together. We update end markets only once a year. So we're working on that and more to come on that probably in our next earnings call.

Okay. Yes, because I just asked the question because you've got a lot of really interesting businesses that I knew that old -- I've been doing this a long time. I know the companies you acquired. You have PCR is in there instead of Microsemi...

Yes, FPGAs -- FPGA and other and all of -- everyone in here is timing on their devices. Microchip keeps the world's time. All of the cesium clocks or every GPS -- every credit card transaction, that's all based on our atomic clocks that are out there. And that's also part of other that's in there.

Okay. Great. I look forward to more clarity on that. Underutilization, you mentioned inventory is still north of 200 days. You've been absorbing underutilization charges. What's the target there? And at what point do you view that as being less of a headwind.

So that's going to take us a while to work through. I think in the December quarter, last reported quarter, our utilization charges were about $51 million. And the majority of that is coming from our 2 large wafer fabs. There's some of it that is in our assembly and test factories.

We built up significant capacity during the last up cycle, and we have to grow back into that over time. We did do a fab rationalization, where we shut down our Arizona fab, and so that's in a better position, but we still have a lot of capacity that we need to grow into.

So it's going to take us some time -- some of the growth areas that we'll talk about are actually not internally produced from a fab perspective. Anything in FPGA, most of Ethernet, our data center products are all externally produced.

And so those growing is not going to help us with the fab utilization. It has to be more of our standard microcontrollers and analog, timing, memory products that grow to help us grow back into that footprint.

And it will happen, but it's going to take some time. I think it's still going to be a couple of years that those charges hang on. Now we'll be increasing production as we move through and revenue is increasing. We're increasing capacity this quarter. But you shouldn't expect that $51 million charge to go down by $15 million or $20 million in a single quarter. It's going to be more gradual.

Okay. And what is the target inventory level? And is that different from the external versus internal fabs?

So our total days of inventory target is between 130 and 150 days. We ended last quarter with still 200 days of inventory. So we have progress that we need to make. We are significantly underproducing from what we're shipping today from our internal fabs. So it is coming down and will continue to come down.

And we can't wait for it to get to 150 days until we start to ramp because there's some limitations on how quickly we can add people and get them trained and start more wafers in the factory. So we'll be doing it gradually and making sure we don't get too low, but we still have a ways to go on inventory correction internally.

Is the reason you don't take it down further quickly is because it ends up being disruptive?

Well, there's a mix issue, right? During the up cycle, we were adding capacity and building inventory at the peak of the cycle. And when the cycle crashed and revenue went down so much, the product that we have was out of mix from where orders are coming in today. So it doesn't mean that product isn't going to sell eventually, but it's going to take time.

Many of these devices, it's a foot deep mile wide application space, right? So there's -- some wafer lots we produce may produce products for 5 years off of that. So it's going to take a while to lead that off.

Okay. Makes a lot of sense. You have talked about 65% being a realistic long-term gross margin target, and that seems consistent with your business mix. But it's going to take a while, as you said. Any other aspects to that improvement besides utilization?

So it's utilization and then it's product mix. So we are guiding to a midpoint this quarter on gross margin on a non-GAAP basis at 61%. So there's 400 basis points improvement to come. If you added back the underutilization charges from last quarter, we'd essentially be there. Again, that's a longer-term process to get there.

But we also have very high gross margin products that are going to be ramping at a faster rate than Microchip is growing overall, whether it's PCIe 6 in data center, the Ethernet products that we talked about in our last earnings call, FPGA, and they all have higher than corporate average gross margins. So that will be a tailwind to gross margin. Again, it doesn't help fab utilization, but helps with the overall profitability.

Makes a lot of sense. Maybe if we could talk about some of the growth drivers. Data center, you've been pretty vocal on PCI Express Gen 6. Can you just give us an overview of that technology? We had [ Adsterra ] on stage earlier today. Obviously, they've had some first-mover advantages there. Can you talk about your progress?

So we launched our product line in September of last year. To date, we've gotten 4 confirmed design wins. One of those is -- one design win is $100 million plus per year usage dollars. There's about 20 others that are in design at different customers. Feedback that we're getting from customers is the power consumption is about 20% to 30% less than everyone else's.

And typical gigabit data center, PCI switching is about 8% of the power, 20% power saving is pretty significant. And then the security that's built in. So all of the form fit and function is really exceeding a lot of customer expectations.

And for short distances, the fidelity of the device is -- the feedback from customers is so good that there is not a need for retimers in some of the applications. And we're also working on a series of retimers to match up with this PCIe Gen 6 product. And and we'll be releasing that over the next month or 2.

I don't think Rich said this, but just a reminder that we are the only company that has a 3-nanometer Gen 6 switch. So we are the most advanced product out there in this area.

That's helpful. And you have a long history in PCI.

Yes. So we started in PCI switches with Gen 3 and Gen 4. We were the market leader in Gen 3 and Gen 4. Unfortunately, for Gen 5, we decided to roll and build our own SerDes inside rather than licensing it. That delayed our release to market by 12 to 18 months, and we lost our #1 market status.

And so -- we decided -- in the midst of that, we decided not to come out with a 5-nanometer device, but to jump ahead to 3-nanometer and take that risk. And then we developed a bunch of IP partners to help develop and bring that vision to life.

And the whole reason to jump to Gen 6 on 3 nanometers was to be able to quickly come out with Gen 7 within 12 months. And so we are planning to have that Gen 7 device within a very short period of time.

That's helpful. In general, as you think about data center, it seems like a lot of opportunities for Microchip, knowing the portfolio, again, of the historic acquisitions that you've done. Can you talk about your focus in data center? Any other things like besides PCI?

So within data center, we've also got timing products, precision timing products that we build around these products. All GPUs utilize our security compute devices for Root of Trust in data center. We also, for AC to DC or DC to DC power supplies, almost all of those power supplies using our microcontrollers with integrated DSPs to form the basis of their digital power within data centers.

And that's a pretty big growth engine for us in that space -- so it's not just the switches, there's many other areas. And then what we're also seeing now is that most data centers are now implementing our cesium-based clocks into the data center for precise clocking within the racks.

I was an engineer, I did clock chips 35 years ago, engine history. FPGA for you guys. I actually -- speaking of me being old, I did the Actel IPO covered it for a decade.

Really?

It didn't really grow until you bought it.

It's actually grown significantly because when...

After you bought it, it grew nicely.

Yes. When it's fascinating, they were so focused on the aerospace and defense market. They failed to see just if you brought that low-power value proposition to industrial control, medical markets, there was so much more opportunity for that.

And -- and so that's essentially what we did is Microchip is our largest market segment is industrial control. And so we brought that into all of our customers, and we've seen tremendous uplift in design wins. And now what we're doing is where we've worked -- one of the things that we had to educate our FPGA team on, we shouldn't -- it shouldn't take a PhD to write the software code for every FPGA.

And I remember going out and visiting with the development team a couple of years ago and I sat down with them and I said, it is ridiculous. I don't want a PhD to write the code. I want it much simpler.

And now we're bringing out a whole development suite later on this year, much like we did with our microcontrollers that use AI and VS code and will generate a great deal of the code. So you don't have to have an advanced degree to get a product out to market.

It seems like a pretty exciting space. I mean, given the 2 biggest companies have been acquired and various distractions that they have to deal with kind of you and Lattice have really sold opportunity to really pursue this.

There's really only 4 players out there within the FPGA space, right? And where we're really winning is in just that low-power space. And so a lot of customers will bring us in for those particular low-power fleets.

We're still winning a great deal within satellites that are being launched, the Artemis rocket that was up there, whether it's an Airbus or Boeing, our FPGAs are all over those particular products.

Yes. Great. Cool. Connectivity, automotive, industrial connectivity, you have a broad portfolio of products in those markets. Can you talk about the progress there? And when does that turn...

More top line growth? So we announced just this past quarter that our working relationship with Hyundai Kia Motors. And so we've designed in our Ethernet products within the Hyundai Kia Motor software framework. We are working with other manufacturers.

We just -- BMW just announced something with ASA, where they're using our connectivity for cameras and advanced vehicle architectures going forward.

And so we've -- Ethernet, whether it's in data centers, industrial control, robotics, a lot of the humanoid or robotics that are being built up out there, all are using our Gen 3 and Gen 4 12-lane switches in those humanoids. And that's also a new growth area connected with our Ethernet products. So a lot of activity there. A lot of this ramp or this conversion starts to really take hold later part of this year, 2027, 2028.

Okay. Okay. Great. And then aerospace defense, a market that is definitely in favor with semiconductor investors at this point. You guys have a pretty big portfolio with the Microsemi assets, big increases in defense budgets. Can you talk about the growth prospects? And then maybe wrap in low earth orbit and data centers in the space, any excited about there.

Yes. So that is a huge growth area for us. Typical satellite may have anywhere from 25 to 200 devices from us inside of it, whether it's FPGAs or power products, discretes, Mill 19, 500 devices or all part of those satellite systems.

So there's -- there really isn't a satellite system that launches without something from Microchip in it, right? And it's just that the LEO launch, and there's significant content from Microchip in that satellite system. And almost anything that goes to Earth's atmosphere has -- from any country has something from Microchip in it.

That's helpful. So maybe moving away from the products, talk about some of the trends. Can you talk about pricing? You've talked about pricing discipline. Are you seeing any change in competitive dynamics there?

You've actually seen a couple of companies raise prices on the back of commodities prices rising input costs. Just how do you think about the role of pricing?

Yes. So pricing has been stable for us. We have not done a wholesale price increase on our customers. We are very conscientious with our customers today about preserving and enhancing relationships. That was one of the 9-point plan points was to get our customer relationships back where they needed to be.

So if there is any area that we see a significant increase in raw material costs or whatever it might be, we would selectively share that with the customer and raise prices on them, but this is not any sort of wholesale program that we have. So no intention to do that.

Okay. So really more stability than...

Yes. I just mentioned before, we have become arrogant in some ways and just trying to stay away from that.

And then China, an important market for you guys, one where you do have some domestic competition, but you've talked about it as being relatively small impact. Can you give us a sense of -- is China different than the rest of the world.

Yes. So China, where we're focused in China, China is really acting at typical with our customers, China speed, right? They're trying to get out the best products as fast as possible. Cost is always part of the conversation. It seems like acting with speed is more important to them right now.

And so that's where we're focused on. So we're bringing in our Ethernet products, our ASA products. And so they're really fascinated by the new technology and how can they use it and get newer products and newer technology out to market sooner.

So that's where we're focused. I think there was too much discussion on China for China and this. And at the end of the day, what we found in meeting with a lot of customers, just give us great technology and a reasonable cost and let's move.

Yes. And do you see any parts of that market as being vulnerable to domestic competition that's different than other parts of the world?

There's always been -- I remember my first acquisition that I did there was always domestic competition of telecom way back when -- where they were building the products locally. It's always been there.

There's always whether someone that has similar products to our microcontrollers or analog products with the same part numbers, that's always been a part of the marketplace there.

Yes. So I think I'd summarize it as we are going to be cautious about it. We're going to make sure we're protecting our intellectual property. That is really key for us that we're not turning over our manufacturing technology IP, so it can be copied. And that's going to be important for us to maintaining our margin structure long term.

Yes. And I don't think building there is as important as making sure that you have the right technology at the time.

And is the traditional microcontroller business changing at all? I mean there's been sort of a migration from a bunch of custom architectures to more ARM types of technologies. It doesn't seem to have changed the margin dynamics at all, which you said it wouldn't, but...

Yes. I think there's a lot of -- I think ARM is still prevalent out in the marketplace, but I think customers are interested also in RISC-V now to some degree, whether it's 64-bit or 32-bit, they're interested in common accelerators between the devices as they adopt AI on the edge to do more autonomous decision-making.

So I think the core becomes less of importance rather -- I think tools and software support are actually more top of mind for customers than what the hardware actually is.

Helpful. So I have a couple more financial questions, and then I'll open it to the audience. Leverage, is it still a little bit elevated? How quickly are you looking to pay down debt? Is there a cash return in the future as you think about that? It would be above and beyond the dividend, which...

Yes. So leverage is still high. Essentially, we've flatlined our dividend at this point in time. The Board is 100% committed to keeping the dividend where it's at, so investors don't need to worry about the dividend being cut, but that's consumed a lot of cash flow, and it was really just last quarter where we had the first quarter in quite some time where we generated enough cash to pay the dividend.

We were borrowing for a period of time. And that's not a good position to be -- so we've turned the corner on that. Obviously, revenue and earnings and EBITDA is growing significantly. So net leverage is coming down, but we're still focused on anything above and beyond dividend is going to be used to pay down debt for, I would say, the foreseeable future.

We haven't set a new net leverage target. We set one back on our Analyst Day in like November of 2021, which was 1.5x. And when we got below that, we were returning 100% of free cash flow between dividend and share buyback.

Unfortunately, at that time, that was at levels of earnings that were not sustainable during the up cycle and then EBITDA crashed and we got ourselves in a situation, where we issued essentially an equity instrument that's mandatory convertible preferred last March to ensure we maintained our investment-grade rating, which is important to us.

But we don't want to find ourselves in that again. So focus is going to be maintain the dividend and anything above and beyond that, pay down debt. And I think we're going to be in that mode for some time.

Yes. Okay. And then just generally, how are you guys feeling about all of this? I mean you talked about that the company had become kind of arrogant about certain things.

You feel like you're on a path now that people are excited.

I know Steve in his book talked about shared pain and kind of we don't lay people off, and I know you had to because it got very difficult. How are you guys feeling about the trajectory from here from the standpoint of morale and customer relationships and things like that?

I think employee morale has definitely improved a great deal, right? And that the whole PSP program created a lot of tension with customers for a period of -- we are beyond that with customer interfaces. We spend a lot of time visiting customers on the road.

Next week, I was invited to do the keynote for opening of embedded world, where I'm talking about autonomous edge and what we're doing on creating AI at the edge for our customers, right, the software environment and everything we're putting together and have a ton of customer meetings all next week, totally full. So customers are graciously and we're graciously working with them to much better than we've ever done in the past.

Great. Well, I have to follow up on AI at the edge. So a really popular theme this time last year that people kind of have moved away from and AI is now just a data center thing for people. But ultimately, when the AI gets deployed, there's a lot of applications where we're going to need stuff at the edge. What are you going to talk about next week?

So in one application, there's -- I have a narrative. We have a drill, a ready battery-operated drill, right? And you wouldn't think that you would use an ML model inside a drill. But we work with the customer, we collect the data. We figured out how to get 15% more battery life from that drill by putting a simple model, about 200,000 bits of memory on that drill.

No hardware changes, just software, 15% more battery life. That's a drill. And so I think as we're working with customers, a lot of applications with just simple software models or changes just -- and we're not talking large language models or Gen AI.

We're just talking ML, machine learning models, right? Well-done suite of models that customers can use for wear and tear or vision, keyword spotting, a number of different applications where you could benefit from this technology.

Yes. That's very cool. All right. So let me stop there and see if we have questions from the audience.

When do you think we'll start to see the impact of these new ML models or edge AI, physical AI? So what would that impact be on your company financially?

Yes. Great question. So we spent a lot of time visiting and working with customers to find out what they -- what do you really need, right? Are we just going to throw out a microcontroller with 100 TOPS accelerator and then go to town on it.

And what we found is that some microcontroller manufacturers have 4 versions of accelerators and 4 different software making it relatively impossible to scale or move this around. So what we did was we formed a group to just focus on the software suite, no hardware.

So -- and that -- we're a hardware company, right? And so we gave this group a different charter. We said, we want to develop the best software suite. And we've been acquiring companies and putting that together and create that software suite.

And then I worked with and the team, CEVA Technologies to bring in CEVA accelerators anywhere from 50 TOPS to 1,000 TOPS, so that we can put that same accelerator across our FPGAs, our risk-based products, our ARM-based products, our 8-bit products or what -- it doesn't matter, right?

But what matters is that we have the same accelerator topology anywhere from 55 nanometers to 3 nanometers, but we have a software suite and model suite that can work across all of the products, right? And so that's the division that we set up for this group.

And that vision is resonating well with customers because in many -- more than half of the applications that we found, you don't need an accelerator. You just use your standard hardware, just help me develop a model that I can program on a device.

Does it have sufficient memory for me to run that? And do I have sufficient memory to store the code required.

What about SaaS model.

Yes. So there will be a SaaS model to that, Software as a Service. There will be a place for people to put models in terms of working with that. So there will be a SaaS portion to that as we go.

SaaS as well as hardware as well as just trying -- like the battery-operated drill, it just was not obvious to customers. They had people coming in and selling, no, you have to buy this new $10 microcontroller with a 100 TOPS accelerator on it. No. Just keep buying the same $1.50 device and add the software.

We need to replicate the cloud in our power drill.

Yes. And there's a lot of applications like that, that we're finding out there, a lot of them.

Great. Well, that brings us up to the end of our time. So Eric, Rich, thanks so much.

Appreciate it.

Thanks, everybody.

Thank you.

Okay. Good morning, everyone. We'll move on to our next presentation. I'm Chris Caso, Wolfe's semiconductor analyst. So thanks for joining us at our conference. Next up is Microchip. With us from Microchip is Matthias Kaestner. Hopefully, I pronounced that correctly.

Matt is good.

Matt is good. That's fine. And Sajid Daudi from Investor Relations. Matthias is the Corporate VP of Auto, Data Center and Networking. I know Microchip has had you out on the road a bit with investors recently because we've heard data center is kind of good right now. So thanks for joining us. I know -- and Eric, if you're listening, I hope you feel better.

So maybe to start and Microchip has been pretty vocal over the last quarter or so with a view that the cycle is really starting to turn. And of course, Microchip was a little later to see the recovery, and now it sounds like that you are. So maybe you can give us an update of kind of what you see from a booking standpoint, from what's going on with customers, particularly as we go into the Chinese New Year holiday, which I know is an important time for you.

Yes. And before we begin, I'll just kind of share the safe harbor statements, which is that during the course of this discussion, we'll be making certain forward-looking projections regarding the future outlook of Microchip, and we refer you to the SEC filings that identify important risks. And so with that, yes, I think what has most significantly changed in the December quarter is the distribution channel has largely normalized.

So as you know, the sell-through, sell-in gaps came down very significantly during the December quarter, which was down to about $12 million for the quarter. So we believe that kind of quantifies that distribution is largely corrected. Additionally, we're seeing pretty strong and interesting demand patterns, right? So during the course of the December quarter, bookings activity was significantly stronger.

We were seeing the book-to-bill was also substantially above 1. We're seeing expedite requests that have accelerated, especially over the past couple of quarters. So really just continued momentum, continued recovery and really seeing this momentum build into the March quarter, which kind of calls out for the guidance that we provided.

And maybe you could separate out -- well, maybe first to start that does it feel that you're now shipping in line with end market growth right now? Or do you feel like you're still undershipping the market and maybe that's probably not the same answer for every end market?

Yes, exactly. I think really, it's difficult to kind of quantify, make an end market statement with 10,000 customers and having the large exposure that we have with distribution side. And so -- but we do believe that it is normalizing every quarter. Like I said, distribution is largely corrected.

On the direct side, we still have customers that are buying certain products to consumption and certain products, they're still burning inventory on. So that element still kind of continues. So it's difficult to comment on because we just don't have that visibility, but our expectation is that things are normalizing across both sides of the...

Largely in line with end market demand today with some isolated pockets where customers [indiscernible]

And I think what's also been interesting in this cycle is there's different end market growth dynamics depend for the different end markets. And again, under your purview, auto and data center, they're very diverging end market trends there. And so as you look out over the last year, data center has been a big contributor here. Maybe you could help to level set us for how big data center is for Microchip and the extent to which that has driven some growth over the last year, even though you're still going through an inventory correction in some of the other markets?

Data center is one of our larger market segments, so is the automotive market as well. Data center grew in share within the overall Microchip revenue portfolio, while automotive had declined a little bit, but also coming back. We do see strong demand from the automotive side as the cycles are normalizing and the carmakers are doing better in general.

So I mean, data center is a hot topic for everybody is firing on all cylinders, driving some of the shortages on the memory side, et cetera, but that's also a driver for us at Microchip. But Microchip is not dominated by the end market segment data center.

And then we report data center at roughly 19% of last year's fiscal year's revenue, and that includes compute and data center. And then as you know, we roll up our end markets once a year. So next quarter, we'll be providing more color, we'll have that granularity. And then there is definitely talk about breaking out and providing additional data. So next quarter, stay tuned, we'll provide an update on the exact.

That will help us to get the -- do the math on exactly how much of a contributor that was to -- just looking forward, you guys have talked about better than seasonal growth for a couple of quarters now. How far out should we expect better than seasonal growth?

And I guess there's 2 factors that would drive that. One is, do the end markets start accelerating, do your customers start to restock? And then where you're still undershipping demand and snapping back. So for how long should we be thinking about better than seasonal growth for Microchip?

Yes. I mean, I think as you kind of look back at some of the commentary that we've made over the past couple of quarters, visibility continues to improve, though 1 quarter out is really where our best look is. And so our outlook for the March quarter is 6.2% growth and which is above seasonal, typically and I'll argue that nothing has been seasonal over the past however many years. But regardless of that, typically, we see 2% to 3% growth in the March quarter, and we're delivering 6.2%. So we're optimistic.

Our visibility beyond the March quarter is still kind of improving. And so really, our expectation is that we'll continue to see solid strength that this momentum continues. But really, beyond the current -- beyond the March quarter, we're really not making any comments other than we expect this momentum to kind of continue.

Right. And with respect to -- I think on the last earnings call, you started to indicate that you did start to see lead times extend for certain products. Maybe you can go into that in a little more detail. How widespread is that? And obviously, as lead times extend, that gives your customers an incentive to give you a little more visibility.

We still have about 200 days of inventory. So when it comes to our internal manufacturing wafers coming out of our own wafer fabs, we don't have any issue. We maintain the short lead times in most of the cases where we see lead times extending for the most advanced foundry nodes that we're using all the way to 3 nanometers and for some the assembly components like substrates that are currently getting tighter and where we experience longer lead times also our suppliers in the [indiscernible].

Right. And yes, lead times have essentially just kind of bounced out bounce around the bottom in that 4- to 8-week time frame. And so for that extension, you probably need tighter capacity and kind of longer lead times expansion.

Right. But point taken, there's a difference between what you produce internally and externally. And for the data center market, which is strong now, I believe the majority of that is external.

That is correct.

So maybe this provides a segue into what you're doing with factory loadings and the impact of margins. And you did indicate that you were planning to increase the factory loadings on this particular quarter. Talk about why that's the case because I believe the internal business is more of sort of classic Microchip more broader market as opposed to data center.

Yes. Our manufacturing footprint today is about 37% to 40% internal, 60% external. On the internal side, anything 110-nanometer and higher is mostly internal products. And like we were talking about the lower 90-nanometer and lower on the external side. And so given today where we're seeing the momentum in aerospace and defense, networking, data center area, those are higher -- richer mix products and so higher-margin products as well.

And all of that activity is happening in the outside foundries. Internally, we reported underutilization, which is really the only headwind to gross margins today after the March quarter, about $51 million is what we reported. Like you said, we will probably expect to see improvement in that, but it will be more at a steady clip rather than a step-up function in any capacity.

If you look at our guidance for the March quarter, which is calling for 61% gross margin, that is in addition to the fact that last quarter had a good licensing revenue benefit, which we kind of contemplated in our original guidance. And so as the momentum continues, we'll see. Now if the momentum is stronger, we may inflect some of this higher, but our expectation is that you probably should see a gradual gross margin improvement, most likely in the line of what we're reporting for the March quarter, about a 500 basis point improvement -- 50 basis point improvement.

50 basis points. Yes. And to be clear, a big part of the improvement that happened in the March quarter was elimination of the inventory reserve charges. That is correct. And so it sounds like that headwind is pretty much gone by the end of this quarter and your incremental benefit is from mix and improving utilization.

Mix improving utilization. And also, if you remember, we've been writing off a lot of inventory. So the benefit of that is -- it's already been expensed to cost of sales. So that will be a good flow-through coming from there. The challenge with that is that there's a lot of variability. You just don't know how much of that you would sell. We know there's no obsolescence risk on that inventory. But the next 3 months or the next 3 years, it could take that long, right? So it's kind of not predictable or forecastable.

Complete inventory is going to be more than 3 months.

Yes, for sure.

Right, right. But what about -- I mean, in terms of load and the impression we've gotten from prior calls is that your gross margin target is 65%. So it sounds like you're still confident in achieving that. It sounds like that, however, based on what you're talking about with regard to factory loadings, that's not over the next kind of 2 or 3 quarters. It's a bit -- we probably think into next calendar year. Is that a reasonable...

Most certainly. And we've kind of publicly stated that, that will come at a slower pace and probably carry through most of this calendar year into next calendar year. So our path to 65% is very much achievable. We are very confident in our ability to achieve that.

Obviously, we've come back from a gross margin low of 52% or something. So we've recovered about 1,000 basis points over the past year as we've implemented our 9-point recovery plan, but still a ways to go. So the path is clear, and we have -- the timing of that remains -- depends on how strong the market...

Maybe we'll pivot to some of the markets and Matt, while you're -- there's been a lot of interest in Microchip's data center exposure. And I think perhaps some of that interest is the fact that people didn't realize that you had data center exposure, which is perhaps one of the reasons why you're here. So maybe talk to us a bit about where you do participate in data center. You mentioned some of the PCI Express wins on the earnings call. And how broad is the portfolio? Where is it positioned? And how much of it is new?

So our data center business has basically 3 main pillars of products. One is PCI Express switching, PCI Express retimers, and we do this for generations. We're not new in this game. We do this for Generation 3, 4, 5. And now we have Generation 6, which is the latest product out in the market. We're working with customers. It's the first Generation 6 PCIe switch built on 3-nanometer technology that provides a quite significant power advantage over other products that are out there. So this is creating quite a lot of interest.

The other pillar is flash controllers for the large flash storage systems, which includes a complete software stack that comes along. So a very complex product, and we get a lot of traction there. Also, this is not a new family of products, but with every generation, basically the speed doubles. And then PCIe Gen 5 to Gen 6 is doubling of the speed per lane, usually also adding more lanes. So that's the name of the game in the data center connectivity world is all about speed.

And then the third line are HDD controllers, which HDD is considered the older technology, but because they have the cheapest cost for storing data, they will have a very, very long lifetime and a lot of the hyperscalers and data center providers are committed to use HDD for their storage needs. When it comes to content like video content, audio content, et cetera, where the highest access speeds are not really required.

So our hard disk drive is good enough. But of course, also this technology is improving, and we continue to develop products for that segment of the market and have good customer traction where we're sampling to customers where we're closely aligned with them to define the features that they value. We add features not only doubling speed from generation to generation, but also adding features like security, post-quantum cryptography is a big trend.

Quantum computers are not there yet to break the cryptography, but everybody wants to be ready so that the deployments they do in data centers are future-proof. Those are all things that we embed into our products today and where we believe we have a very good position.

Right. So -- and maybe to start with the PCI Express Gen 6. Maybe talk about for some that may not know, the application for that. So why is PCI Gen 6 important?

So within a data center, you got the main processor, the GPU, the likes like NVIDIA, we're not providing this type of product, but a stand-alone GPU doesn't work. It needs to be connected to other GPUs. It needs to be connected to high-speed memory. It needs to be connected to the outside world. And for that, a lot of switching technology and connectivity technology is required. And PCI Express is the de facto standard interface on each and every SoC of every GPU, SoC for servers, and that's where PCI Express is coming into play.

In the past, the cycles from generation to generation were 3 to 4 years. This because of the need in the data centers to have higher connectivity speeds has been compressed down to 18 months to 24 months. So we beefed up our development team and effort to be able to work with this cycle and to be on the forefront and having the first competitive products out with each new technology cycle. The number of PCIe switches in data centers varies depending on the architecture, depending on the GPUs that are being used, but a ratio of 3 to 4 to 5 high link count switches per GPU is not uncommon. It's a substantial number and a substantial market for us.

Right. Is Microchip participating in UALink as well?

We do have activities on the Ethernet side, but UALink is one of many standards that is moving forward. So if you look beyond PCI Express Gen 7, there are competing technologies. NVIDIA is proposing their own evolutions of NVLink to cover some of those markets as they're doing it already today. So this is not new. But there are some more proprietary solutions that are also pitched by some of the industry players in that space. But UALink is certainly on our radar screen.

Maybe speak of the FPGA business as well because that has been another -- I don't know, maybe I say surprise grower, but probably not a surprise to you with good margins. So where is FPGA fitting into the picture?

Our FPGA business is growing very strongly at the moment. So it's one of the ones that are driving our revenue forward. And it's across many different markets. I think the original strength was in the aerospace and defense segment, where we have also radiation hardened product for satellite operations, et cetera. It's a high-margin business and the aerospace and defense market segment of all of our segments is probably among the strongest today for the obvious geopolitical reasons as well. But the FPGA business is not limited to that.

There are good applications on the industrial side, for example, industrial robots. There are applications on the automotive side, and there are also applications on the data center side. In the past -- and that's advanced process technologies, and ASIC development wasn't that expensive. So it was justifiable to develop a custom product for a specific application, even though the volume is not extremely high.

Now with the mass cost and development costs going into triple-digit million dollar figures for the most advanced nodes, you need significant volume to justify an ASIC development. And that's where FPGAs come into play. They use when customers have a specific need to translate protocols to process video data and things like that. So this is across many different industry segments, including data center, including industrial, including automotive and of course, aerospace and defense.

And just one point of clarity just for the audience. I think as we report from a reporting hierarchy standpoint, we report an MCU category and analog category and other FPGA largely sits in the other category as people kind of look at our line items and see just as a highlight for folks.

Right. Then that was an area of strength in the last quarter, but part of that was from licensing, but then FPGA.

Then FPGA was a piece of that as well. And some of the feedback that I got post earnings was that people thought we missed on the MCU analog side where overall, like I said, we contemplated the licensing piece in our original guidance and the upside came from the silicon business, which December quarter, again, seasonally typically is down 2% to 5%. So there's some kind of culture confusion. So hopefully that clears.

I'll repeat the question for the webcast. It's just the question of PCI Express scale-up versus scale-out opportunity -- PCI Express switch scale-up versus scale-out opportunities.

We do see it in both areas. And it depends a lot on the data center architecture that customers are choosing, but we do see deployment of PCI switches and scale up as well as scale out.

Pivot over to auto a bit. Maybe you could speak about the microchip exposure in auto. Historically, the microcontroller business has had a strong auto component. And I think people understand that. Maybe you could talk about some of the other areas you're participating in auto as well, which may not be as well understood.

So historically, microcontrollers was one strong component, but we were, I think, always strong in specific applications. For example, touchscreen controllers or touch controllers in general for touch buttons, surfaces, lighters all the way to the touchscreen where we have a significant share of the market. Car access is another element. So the chip that goes into the key fault, but also all the electronics that are inside the Car to realize the hands-free functionality, approach the car and pull the handler open without taking the key out of your pocket to make this in a secure manner.

It's a tricky thing. So a lot of electronics involved, and we are also one of the 2 big players in this area. It's not just microcontroller itself, complete solutions that go around. We are historically also working on the in-vehicle networking side. So we have connectivity products for LIN, for CAN, for most and for the very early Ethernet connectivity as well in automotive. And we do see a trend that today that up to 20 different technologies used to transmit data throughout the vehicle, which is a challenge to scale.

It's a challenge for the software development. And the big trend is that people are moving away from the multitude of different standards towards a unified Ethernet-based system. And we're extremely well positioned in that as we develop switches for automotive, develop dedicated transceivers from 10 megabit, 10-based T1S all the way up to gigabit speeds that are used in automotive.

While Ethernet -- the Ethernet itself will reduce the software complexity, will reduce the development times of cars with that. So that's important for the carmakers. In addition to Ethernet, we believe that there will be 2 other pillars that will be embedded in future vehicles. One is PCI Express. We have a main computer in the car doing all the ADAS functionality, all the in-vehicle infotainment functionality. And you have high-performance processors.

They need to be connected to each other, also functional safety reasons. They need to access a common high-speed memory. So for that PCIe switches are also required in automotive vehicles, not the same class of switches with the same number of lanes and the same speeds as used in data centers. But for automotive, we derived an automotive switch from our PCI Gen 4 data center switch, added the safety and security features to it as required in automotive, cut it down from the number of lanes from a cost point of view to make it suitable for automotive use. Without being in data center, this would not have been possible because the initial market for that is relatively small. But because we could derive this technology from our data center products, we had a unique opportunity to do so.

And we expect this market to grow fast. By the way, not only in automotive, the same product, we got a lot of interest from server makers, from industrial, for robots, et cetera. So this is proliferating beyond automotive. The third element in automotive, where we develop connectivity solutions and made an acquisition in April last year is the connectivity between the camera and the main computer and between the main computer and the displays.

Today, proprietary technology from competition is used, but the industry is driving towards open standards. And we teamed up with some automotive companies, some OEMs, including BMW, to develop a standard called ASA, Automotive Serdes Alliance. And we have the first products out in the market, get a lot of traction, a lot of proof of concepts, et cetera. Automotive is a slow-moving market. But once you're in, you're in there for 7 years, 10 years for certain vehicles.

So the traction there is tremendous on both ASA, Ethernet as well as PCIe. So I do believe that the future growth in automotive, a significant portion of the future growth in automotive will come from the connectivity side. And just to give you a perspective, CAN, controller area network is the legacy standard that's in cars for more than 25 years now. There are more than 2 billion CAN nodes a year deployed in cars. And with the new Ethernet 10 megabit standard, 10-based T1S, we have a chance to replace many of them over time.

So the market potential for that is tremendous. The benefits for the customers are tremendous. We have everything on the same standard so that the upper layers of the software stack remain the same, while when they have different need for different speed grades, they just exchange the transceiver layer, the lowest layer of the protocol while they can keep the software. Today, they need higher speed. They have to redo the complete system. That's a nightmare.

Now I'll add one just like a dynamic to it that I've been seeing evolve. And this is more of a very long-term kind of statement here. But historically, as you guys have heard about Microchip technology, we talk about our anchor lanes being microcontroller, where the engineer starts their activity there, and then we can build around that with our analog products and so on and so forth.

Most recently, we're now starting to see a couple of new pillars emerge, right, one in the data center and the connectivity piece, one in the compute side and AI/ML, where conversations with customers are starting with that as the core application and then spreading into other attach areas to say, well, how can microcontrollers help with that and so on and so forth. So I think that's a different dynamic that historically, we haven't seen at Microchip, and we're starting to see that emerge. It will play out over time, but I think it's definitely real and happening as we kind of think about the future of Microchip.

I mean maybe it's a segue into the microcontroller market itself, too. And one of the trends that you folks have called out post pandemic has been the migration finally to 32-bit. And for years and years of covering Microchip, I've been corrected oftentimes by certain Microchip people about why there are certain applications, which are just fine for 8-bit will stay with 8-bit. Are we seeing that migration? I know that migration is happening more broadly now. We're seeing that migration happening within the auto market as well.

I think if you look at microcontrollers, it's not so much about the core anymore. And the more advanced process technology nodes, the cost of silicon transistor has come down quite a bit. So it doesn't really matter whether it's an 8-bit [indiscernible] Customers also really don't care that much about it because the microcontroller, including all the analog peripherals, needs to have the right interfaces, the right HDD converters, et cetera, et cetera. But because the cost of transistors came down such so much, it doesn't make a big difference anymore, whether it's a low-end 32-bit core like or an 8-bit core. That's why we do the transition to be more in a uniform architecture to help our customers support that.

We'll continue to support and sell our 8-bit microcontrollers for many, many years. There's no plan to life them, et cetera. And there are many customers that are happy with them and use them continue to use them also new applications. But going forward, the bulk of the development activities are on the 32-bit side.

But does it matter to Microchip if -- and understand the core of it's an 8-bit core, a 32-bit core probably doesn't mean as much to the customer. The cost isn't a lot different now. But the underlying architecture is different. Your 8-bit core is proprietary PIC where 32-bit is going more towards ARM. So how do you deal with that? And does it affect Microchip's historic competitive advantage by that migration occurring?

I wouldn't say so because in fact, the Microchip architecture, you mentioned the PIC is not the only one. I mean through the Atmel acquisition, I personally joined more than 10 years ago from Atmel side, with the AVR architecture. So we have already 2 architectures. And I think we're pretty well versed how to integrate both architectures and now we are already at ARM cores.

It's not new for us that we're using ARM cores into our tool suite that we give to our customers to develop their solutions using our microcontrollers and our tools. So from that, it's not such a big change because it's not new to us. It's just a shift where we put the focus, but I don't expect that our competitive advantage situation will change because of that.

We've got about a little under 10 minutes left. Are there any other questions from the audience because I can keep going, raise hand. Anyway, I'll keep going. Maybe we can talk about China a bit. And China market has been volatile to say the least. Maybe you can speak about what you're seeing in China now and how your approach to the China market may have changed as a result of tariffs and some of the trade uncertainty now.

Think China, it is a volatile market. They're trying to build their own semiconductor industry and pushing that. But we learned that most Chinese customers are pragmatic. They're happy to buy our products, even where a U.S.-based semiconductor company if the price is okay and if the features are the ones they need. There is competition, especially on the low-end side, on the microcontroller side, some of the analog side.

But we have tremendous design momentum on the automotive side, where features -- complex features like the [indiscernible] system are not available from local competition. This may come and evolve over time, but we don't have big restrictions in selling into China. Also, the total of our revenue that stays within China is not getting reexported out is relatively low. It's a high single digit. That's it. So our exposure to the local Chinese market is not as big as some may think.

We do see some restrictions for the more advanced products that go to companies who are closer to the government related to infrastructure like the Chinese telecoms. They get told by the government, you cannot buy foreign semiconductor, you need to use local one. There are a few products where the U.S. administration selling you cannot sell to customer A, B or C in China. But the impact on our business hasn't been dramatic because of that. So it's a pragmatic approach that we're taking to China. We keep our sales momentum.

We have good design interaction in China with Chinese local companies. it depends a lot of interpretation there. Those perceived rules, the Chinese customers or Chinese companies need to have a minimum local content. I think it's more pronounced with those who are closer to the government or partially owned by the government. So they have some preemptive opinions. It's not unique to Microchip. So they feel they need to buy Chinese not to lose subsidies in the future. But it's not a major, major impact to our business today.

And we break down China as it's roughly 18% of total revenue and roughly half of it goes into China for manufacturing and then gets shipped out globally. The other half of 9% or so is more complex Microchip content, so difficult to replicate. So really kind of leaving with a sub-5% exposure that potentially over the long term is at risk. But we'll continue to innovate with new features and...

And have you seen any changes? I mean, obviously, the -- from a trade standpoint, things have changed. But from -- you've always competed with China local competition. Has there been any appreciable change in the capabilities of the local companies of the scale that you didn't see 5, 10 years ago?

Yes, absolutely. I mean not only for our product range, but if you look at advanced SoCs, companies like Verizon, for example, in Beijing, others, there are viable competitors to the large SoC vendors. So they increased their capabilities quite a bit in the Chinese government and the authorities they're pushing. So many, many start-up companies doing semiconductors in China, that we all have to deal with, right? It's not unique to Microchip.

World going forward, yes.