All right, let's go ahead and get started. Good afternoon. Welcome to JPMorgan's 53rd Annual Technology, Media, and Communications Conference. My name is Harlan Sur. I'm the semiconductor and semiconductor capital equipment analyst for the firm. Very pleased to have Eric Bjornholt, Chief Financial Officer of Microchip, here with us today. Microchip, top -five microcontroller supplier globally, number -three market share in industrial market MCUs. 60% of the MCU portfolio is 32-bit and higher architectures, solid analog and FPGA portfolio. It's been a busy earnings season, so I've asked Eric to start us off with a summary of the March quarter and June quarter outlook, and then we'll go ahead and kick off the Q&A. Eric, thank you very much for joining us today.
Thanks, Harlan. Thanks for having me. Before I start, during the course of this discussion, I will be making certain financial projections and forward-looking statements that involve risks, and I refer you to our filings with the SEC that identify important risk factors about the company. We've been in kind of an extended downturn, cyclical downturn, for quite some time at Microchip. Finally, we're able to call the bottom of that with the results for the March quarter, which was our fiscal year ended for fiscal 2025, and are guiding positively for the June quarter. We've made quite a few changes. Steve Sanghi is back at CEO, and he's implemented this nine-point plan that he went through with investors on March 3, and we're making progress along those trends. Anyway, we are guiding the March quarter to be up about 7.7% sequentially at the midpoint.
We have bottomed in terms of gross margin. Our inventory is coming down. Our inventory went from 266 days at the end of December to 251 at the end of March and is going down more materially as we head into the June quarter. Just about 10 days ago, we finished the last production at our Fab Two facility in Arizona, which is being shut down, and that is really going to propel our inventory reduction program. We said on our earnings call last week that we expect the dollar amount of inventory on our balance sheet to reduce by about $350 million over the course of this current fiscal year. Looking forward to that. With that, we should be able to start making some really nice improvement towards our new long-term operating model, which is a 65% non-GAAP gross margin and 40% non-GAAP operating margin.
With that, I'll turn it over to Harlan for questions.
Yeah, thanks again, Eric. That was a great overview. I'm going to start off with near-term dynamics. We'll talk a little bit about strategy, and then we'll talk about products. From a near-term perspective, the Microchip team and some of your peers in the analog and embedded microcontroller space have talked about the current trends that point to the early innings of a cyclical recovery off of the downturn that started in 2022. In the March quarter for Microchip, significantly higher bookings. Overall book-to-bill was greater than one for the first time in almost 3 years. Order strength has continued thus far into this quarter. Many would argue that this uptick in orders and near-term demand activity is no different than a demand step up due to pull-ins ahead of the expiration of the 90-day reciprocal tariff reprieve.
We've actually heard from a few large OEMs that they are indeed pulling in component inventories near term. Some of the more, maybe not apples to apples, but maybe some of the more compute semiconductor-focused names like Intel and some of the memory companies have also talked about seeing tariff-related pull-ins. Maybe, Eric, take us through the process by which the Microchip team goes through to decipher the recent strength and distinguish between pull-in demand from potential cyclical-related demand improvements.
Okay. All right. We started seeing these trends in improving bookings in the month of January, and those continued January, February, March, April, and now into May. It has been a continuous trend. We have gone through this massive inventory correction at our customers and distributors over the last 18 months. We are finally at the point where distributors and customers are needing to start to reorder again. We are getting that feedback from them. We can see specifically in the distribution channel that inventory has come down dramatically. Just this last quarter, distribution sell-out was $103 million higher than distribution sell-in. If you look at the last five quarters combined, that number is like $525 million. Distribution is getting at the point where inventory is getting low enough on a lot of their SKUs where they are just having to reorder. Our direct customers are seeing that too.
Now, we've visited with a lot of customers. We've been meeting with customers, and the question is always directed to them, "Are you doing anything different because of tariffs?" The answer has consistently been no, that they're kind of waiting for things to settle out. We do not believe that we've really had any significant upside in bookings activity from tariff-related activities.
If I rewind back to the 2018-2019 U.S.-China tariff and trade conflict, the Microchip team was actually the first to call out seeing order patterns dropping within your China industrial and manufacturing customers due to the demand uncertainty that was caused by tariff impact at that time. The Microchip team has always had somewhat of a unique profile of having large exposure to small and medium-sized manufacturers, and they tend to modulate their activity fairly quickly and in real time. Given the current tariff and trade dynamics, we've seen a pretty steep drop-off in China manufacturing activity. Are you starting to see any pullbacks or perturbations in order activity from your small to medium-sized business customers in the China manufacturing and industrial space?
We really aren't, but I think that kind of leads back to my comment earlier on this large inventory correction that we've been through. You mentioned that we hadn't seen a positive book-to-bill ratio in almost three years. Customers have been in this inventory depletion mode. I think with that, it's probably hard for us to pick up what is happening specifically in China with these small to mid-sized customers because they've all taken inventory down so much and their order patterns weren't regular leading into this. Now when we're seeing the uptick, it's not visible to us.
I want to get a sense of the breadth of the recovery you're seeing. You talked a little bit about this. As of the December quarter, your direct business was about 43% of total sales. Distribution was 54%. You've got very good visibility into your distributors. You can monitor sell-through trends. Historically, you've talked about less visibility on inventory at your direct customers. But you can probably gauge the level of their inventories by looking at their order patterns. I mean, you've clearly observed the increase in sell-through and order trends from your distributors. Have you seen a similar pickup in order trends at direct customers, suggesting maybe inventories for them have bottomed as well, or is that trend still lagging your disty set of customers?
Okay. So maybe just one correction real quick. Distribution is about 45% of our business, not 54%. Anyway, it is a significant piece of the business. You are right with distribution. We get inventory reports every month. We know exactly by SKU what they hold. With their direct customers, we do not get that. We have seen the uptick in bookings activity and order activity from direct customers also. We absolutely believe that not just distribution will grow in revenue this quarter, but our direct customers will also, and those order trends are continuing out into future quarters.
In addition to the strong rise in orders in the March quarter, typically like in a cyclical recovery, you also start to see stronger turns business, orders and shipments that are fulfilled in the same quarter. Did you see your percentage turns business rise strongly in the March quarter, and is that still kind of sustaining here in the June quarter?
Yeah, it is. We have seen a high number of turns. Turns are increasing in the business. Lead times are really short with inventory being high, and customers know they can get product in short order. Now that they are starting to replenish inventory, the percentage of turns that we're getting is absolutely increasing.
Let's talk a little bit about the pricing environment for this year. The team on earnings last week was asked about pricing. It looks like the team is forecasting a mid-single-digit % decline on average in pricing, which is a bit larger than the low single-digit % decline of some of your peers and what they're articulating. Maybe if you could just elaborate on whether this is part of your pricing strategy, maybe to potentially increase market share, or are you seeing increasing competitive dynamics, or maybe it's just a mixed impact, but more importantly, is the team driving its average fully packaged and tested chip costs at or greater than 5% per year as well?
Okay. From an ASP perspective, ongoing business is pretty stable. At the point of new design, we are being aggressive on pricing. We are leading with our newest, best products and going up against all the competition and want to lead with an aggressive price to make sure that we are gaining and not losing market share. I think that is really the dynamic. We came off a couple of years where pricing was increasing. We were getting all these increases in cost, supply chain costs that we were passing on to customers. Now we are at the point with having high capacity, having high inventory. We want to be aggressive. We want to be market share leaders. For the most part, pricing is stable, but on new designs, we are being aggressive.
How do we think about your overall, when you think about your supply chain and manufacturing cost infrastructure, any ways we can think about how you think about annualized cost declines?
We think that this is kind of a short-term thing where we're going to be super aggressive on pricing on new designs. It's always going to be competitive at the point of design, but we're coming off a period that was high growth followed by this big decline in inventory correction. Again, we want to make sure that our customers feel supported. Our cost structures are actually in really good shape. We've taken out this Arizona fab. We think that that is going to take out about $90 million in annual cash savings to the bottom line. Our manufacturing infrastructure was built up pretty heavily during the last upturn. We've got a lot of equipment that's already paid for that we can deploy. Our facilities are in a really good position as revenue starts to grow again and volumes grow to produce outstanding costs.
Yeah. Okay. You mentioned the nine-point recovery plan that was laid out in March. There was a particular part of that that revolved around your channel or distribution strategy. There you aim to improve your channel strategy by lowering fulfillment margins, reducing incentives, aligning more closely with industry peers to enhance product and price competitiveness. You mentioned that you've just completed, or Steve mentioned that you just completed this initiative without any negative impacts. What's the channel partner feedback been regarding these changes? Have you observed any notable differences in either sales velocity or product mix?
We have really good relationships with our channel partners. We sell through over 120 distributors worldwide, the global distributors that everybody's familiar with. We have regional distributors that are kind of country or area specific, tend to only carry a single microcontroller line as an example, are fully trained like our Salesforce and go out and create demand with their field application engineering teams. We also have the catalog distributors. After we made these changes, they were really accepted very well by the distributors. They clearly knew that we were paying kind of above standard market for both fulfillment and demand creation. By tweaking those things down, we have more in line with the rest of the industry, but we are still incentivizing them probably better than the majority of the industry. They want to be engaged with us. They want to work with us.
They love our product line. It is a big percentage of our sales, as I mentioned before. It is 45 percent. It is an important aspect of our go-to-market strategy.
Let's talk about the product strategy. This particular part of the strategy has been around for a number of years. Your total solution strategy, or what we call TSS, and the idea is to drive higher attach rates of Microchip products to your Anchor products, which are typically your microcontrollers, microprocessors, or your FPGA products, more analog power management, networking, chip attach per MCU or MPU design. In the March business update that Steve gave to us, he presented a pretty insightful chart on Anchor products showing an attach rate on average about 4 to 5 additional Microchip products per Anchor product. What programs has the team put in place, sales, marketing, channel partners to drive the increase in higher attach rates over the past number of years?
Yeah. One of the starting points for that is having a robust amount of reference designs based on what the customers are doing. By having that, we can present examples of things that we have done in the past that can bring them to market quickly. It can be using a single semiconductor supplier to support their system. These reference designs, we go out and we not just train our Salesforce through our business units, but we also train the customers directly. We train our channel partners so they can be effective in selling it. The momentum there is continuing to grow where the number of parts per system is an increasing metric that we see as we go through each quarter and every year that it's building.
It's important to lead with the Anchor product, but then surround that with the other products we have in portfolio, whether they be general-purpose microcontrollers, analog products, timing, security, memory, etc.
Are there incentives to sales and marketing to drive higher attach rates to the Anchor products?
Internally, we have everybody on the same type of bonus program. We are a non-commissioned Salesforce. We have everybody focused on the overall revenue growth, profitability of the company. Our distributors have their own incentive programs that they drive their sales teams with.
The team is well known for its leadership in microcontrollers, but over the years, you've doubled down on embedded microprocessors as well. You introduced your 32-bit microprocessor a few years back, strong TSS attach of around 5 additional components per MPU design. You just added your new 64-bit microprocessor family to your Anchor product list. Your 32-bit MPU, again, around for a few years, is based on the ARM Compute Architecture industry standard compute architecture. I noticed that your new PIC64 MPUs are based on a RISC-V compute architecture. RISC-V is gaining traction in the market, but ARM continues to be the dominant architecture. What drove Microchip's decision to move to RISC-V for its 64-bit architecture?
With some of these more complex solutions, customers need more flexibility. RISC-V is providing that for us in this particular area. We are still great partners with ARM, and they do a lot of good things for us. For this particular product, it made sense to use RISC-V. It provides flexibility to the customers, and we can actually see this as an emerging opportunity in some of the edge compute areas that we are focused on.
Is there any plans? Because ARM is well proliferated across your microcontroller product line, your 32-bit MPUs. I assume that there's probably some roadmap at some point to have 64-bit-based MPUs based on ARM architecture. Is that fair?
I don't know the answer to that, Harlan, but the first 64-bit that we introduced was a 64-bit microprocessor, as you said, and we're following that up with 64-bit microcontrollers. I can double-check on that, but those probably are going to be ARM-based. I will double-check.
Yeah, because the predominant part of your MCU family is already ARM-based. That makes sense. Gartner, market share numbers out for calendar 2024. As I mentioned before, Microchip, fifth largest microcontroller semiconductor supplier globally, 60% of your portfolio, 32-bit and higher products. I know it's particularly noisy given the undershipment trends during this downturn, but in 8-bit market, you have a strong number one market share position. You're two times larger than your second, the number two player in this space. According to Gartner, I mean, you did slightly underperform the overall 8-bit market last year. It seems that the Asian, especially China-based 8-bit MCU vendors are performing well in this area. Maybe you can discuss the competitive dynamics you're observing in the MCU market, especially in China.
Yeah. It is a competitive market. I think when you look at market share last year, Microchip during the upcycle peaked at higher revenue, and then we trough lower. I think looking at 2024 is kind of more driven by inventory correction than anything specific with losing market share in a competitive dynamic. China continues to invest in semiconductors, and 8-bit is a part of that. We do see some increased competition there. Where Microchip really thrives is in areas like industrial and automotive where the customers are really looking for a platform. In that platform, they want to have the hierarchy of products that they can move to because most of the time when they start a design, what they go to market with is something different because they get market forces that say, "Hey, I need to be lower memory, lower power.
I need more compute opportunity. I need connectivity. I need graphical display. The China competition does not really give them those options today. If they run out of headroom, then they have to start over and pick a different supplier and buy another tool, and it really slows them down. We see competition on the consumer side, and that is really not the focus of our business. Today, our industrial, automotive, and more complex clients really like what Microchip and some of our other Western competitors bring to the table in terms of depth of the portfolio.
On the team's six megatrend focus, in addition to TSS, you have a focus on six megatrends: edge and IoT compute, data center, AI, sustainability, electrification, and networking outside of data center. From fiscal 2021 to fiscal 2024, megatrend revenues grew about 26% on an annualized basis. That's 2x the growth of your overall business, represented 47% of your revenues in fiscal 2024. If you can just maybe chew us up, what was the performance of the megatrend business like versus total sales in fiscal 2025? And what was the percentage mix of the megatrend revenues?
Yeah. The megatrends, and we posted all this on our website, but the overall percentages did not change that much in fiscal 2025. Our revenue was down significantly through this inventory correction. I think the inventory correction, it did not matter what end market, if you were a megatrend or not, you had inventory that needed to be corrected, and it was. Our view on these megatrends is that they can grow at twice the rate of Microchip over time. They have done that historically. We think that uptick happens again starting this year, and we are super excited about it. That is where a lot of resources are being invested. That is where a lot of our TSS focus is also focused on penetrating these markets, bringing products to market that our customers need to make them successful, and we will grow with that.
Okay. That seems fair because, I mean, that's what you've been growing at over the past four years is at a 2x growth rate. That makes a lot of sense. Within the AI and machine learning area of focus, I think you guys have characterized it as a billion-dollar time opportunity. Help us understand where the team is seeing strong traction with your microprocessor, microcontroller, FPGA products in AI workloads.
Yeah. We are seeing the need in systems where there are many sensors and vision is being deployed, that things like robotics, industry 4.0, where we are seeing a lot of acceleration in these areas for us. That is growing quite rapidly for us. Also, on edge nodes where customers are trying to see, predict when a motor is going to fail or when a battery needs to be replaced, these things are where we are seeing a lot of momentum here.
Based on the fiscal 2025 revenue by end market disclosure you talked about, it appears that most of your end markets declined roughly in line with your overall business, except for your aerospace and defense segment, which actually held up relatively well. As a percent of sales, I think it grew by 400-500 basis points. I assume a significant portion of your aerospace and defense revenues is driven by your radiation harden, what we call rad hard in the industry, rad hard FPGA products, a segment where Microchip has a very strong number one share position. I do recall at the time that you guys did the Microsemi acquisition, I do recall that the team was trying to drive some synergies with the rad hard FPGA portfolio to develop a rad hard aerospace and defense portfolio within your microcontroller products as well. Is that the case?
How successful or not has that initiative been?
That is absolutely the case, and it has been successful. Maybe just backing up to the first point, our aerospace and defense business grew significantly as a percentage of sales in fiscal 2025 that we just completed. The business was down, but it was quite stable compared to the rest of the business. It grew from 11% of total revenue to 18%. Now, we do not expect those percentage trends to continue, but we see a lot of opportunity for that business to do really well. One is the products that came to us through the Microsemi heritage, that acquisition that we did about six years ago. We are also selling around those products other microcontroller, analog, and Microchip-based products that are now radiation tolerant that can penetrate these markets. The US defense budget is supposed to be the largest that it ever has been.
We're seeing the NATO countries also expanding their budgets based on some of the things that are kind of being forced upon them by our leadership here in the U.S. Taking those expanded budgets, having the penetration we always have, and then expanding that to the rest of the product portfolio, we see as a huge growth driver for us.
Let's talk about the manufacturing footprint and consolidation as well as future expansion potential. You mentioned you closed your Fab 2 manufacturing plant in Tempe recently. I know this data is probably a few years old, but last I saw the data in 2021, Fab 2 accounted for about 25% of your total 8-inch equivalent internal wafer capacity. Is that about right? Was that about the right mix? About a quarter?
We haven't broken that out, but Fab 2 was absolutely the smallest of our three large fabs in the U.S. Fab 4 in Oregon is the largest, and then Fab 5 in Colorado is the second largest. Fab 2, which we've now closed, was the smallest. It was space constrained. It was maxed out. We'd had that facility since 1994. If we had to close a facility, that's the one that made the most sense. We really haven't taken away a significant amount of capacity. All the products that we're running in Fab 2 are going to run in either Fab 4 or Fab 5, in other factories. None of that is going to be outsourced. We expanded the clean room capacity in Fab 4 and Fab 5 dramatically during the upcycle. We have room to grow into.
We've got about $400 million or so of equipment that has been purchased and not deployed for manufacturing yet. We've got that to grow into. We feel very good about what our two remaining fabs can do for us from a growth perspective. Just a reminder that we only do about 37% of our production internally. The rest is outsourced to the professional foundries.
Yeah. We've talked to a few investors that are concerned that maybe when the demand environment does recover, the team may be caught short on internal wafer supply. You mentioned you have some room for capacity expansion for Fab 4 and Fab 5. I also remember that your Fab 5 in Colorado Springs is still primarily 6-inch. The good news is that if and when the time comes, not only do you have footprint expansion capability, but you can also drive incremental output by converting a big part of Fab 5 to 8-inch. Is that correct?
That is accurate. Fab 5 came to us through the Atmel acquisition as a 6-inch fab, but absolutely has capabilities for conversion to 8-inch over time. That obviously we have clean room space that then we can expand into and increase output.
With the closure of Fab 2, and you said about 37% of your internal wafer capacity, 37% of your total wafer needs are internal. Is that the right mix? Is that how you guys are thinking about is kind of the right mix for the company as you think about the growth profile over the next, call it, three to five years?
Yeah, I think it stays in that range, 35%-40% internal fab. I do not think it changes much. Maybe some of our more advanced technologies and foundries grow a little bit quicker, but these are modest changes year over year.
Okay. Let's look at that. Okay. So in December, this is talking more about the financials now, the team articulated total underutilization charges of $43 million, inventory reserve charges of $82 million, total of $125 million coming out of COGS. That's 10% or added to COGS. That's 10 percentage points of gross margin impact. These charges sustained in the March quarter. They're sustaining here in the June quarter, implying that utilizations are staying roughly flattish through this quarter. Now that inventories are coming down aggressively, bookings, backlog improving. Steve said that balance sheet inventories, like you said, should be below 200 days in the September quarter. He also said that the team will start to increase utilizations when inventories fall somewhere between 150-200 days.
Is it fair to assume that if the recovery trajectory evolves according to your base case scenario, that we should see utilizations rising sometime in the September quarter or December quarter?
That would be our expectation. Today, we are producing product at a rate that is significantly lower than we're shipping. Inventory is coming down. Our long-term target is 130-150 days of inventory. You can't wait till it gets to 150 to start ramping when there's this big gap between output and consumption. We'll monitor that. We look at our manufacturing activities very closely and our inventory levels. It is likely that sometime in the December quarter we would have to increase output because inventory would fall below 200 days, maybe by the end of September. Again, all these are estimates and can change, but that would be our expectation.
I think maybe even more importantly, when you look at those large charges for inventory reserves and the capacity underutilization charges, which will come away more gradually, the inventory reserves are going to fall dramatically. That should start to happen relatively soon, not this quarter. They were $80 million-plus in the December quarter. We did not break them out specifically for March, but they are in similar ranges. We expect it to be similar for the June quarter. We measure this. We take a snapshot of inventory near the end of the quarter. We look at the last 12 months of sales history and come up with an inventory reserve calculation for accounting purposes that says, "Hey, you should write off X amount of inventory." Now, we have two things working in our favor.
One is that inventory is coming down rapidly, a $350 million-plus projected reduction in fiscal 2026. Sales are growing again. The basis for that calculation is going to spit out a much lower number. That immediately improves gross margin when those charges go away.
That's kind of interesting because you said trailing 12 months, right? If I look at it on a trailing 12 months, wouldn't your revenue still be kind of coming down?
It is. That is why even though revenue is increasing this quarter and inventory is coming down, we are still expecting a large surge this quarter. That is going to reverse eventually. Obviously, we are bringing inventory down so much. There is just a smaller pool to select from. The last thing that will benefit us, I just do not know the timing, is we have been writing off large amounts of inventory every quarter for the last five-plus quarters. Eventually, we are going to get orders for that product. We do not have visibility for it today. Our products have useful lives of 10, 15, 20 years. That product will likely sell eventually. That will be a tailwind to gross margin also.
The way to think about it is that the inventory reserves, $80 million-plus per quarter now, you're saying that sometime between now and whenever you start to increase utilizations, potentially in December quarter, those are actually going to start to drop. Those are going to start to drop first. Is that kind of what you're saying? Potentially?
The first benefit that we're going to see in gross margin prior to underutilization charges going away significantly is going to be a reduction in these reserve charges each quarter.
I see. And then how do we think about starting in, let's assume that in the base case, you start to increase utilizations in the December quarter. Is there a formulaic way to think about as your utilizations increase, we should see X amount of underutilization charges coming down?
We have not provided specific guidance on that to the street. I think the street would like to say, "Hey, at what level of revenue do you project that those underutilization charges go away?" and then they can just kind of straight line it. It is going to be a bit more lumpy than that. We have not provided that guidance. We will keep providing guidance on a quarterly basis on margin. We have two large factors there that are going to propel margin towards our 65% target of taking out the underutilization and the inventory reserves. Those two things combined alone.
Get you to your target margin.
That's our target model.
Yeah, exactly. Okay. I appreciate your participation in our conference today, Eric. Thank you very much. Look forward to monitoring the progress of the team this year.
Great. Thanks, Harlan.
Thank you.