Microchip Technology Incorporated (MCHP) Earnings Call Transcript & Summary

Great. Thank you. Welcome back, everybody. It's my pleasure to have Steve Sanghi and Eric Bjornholt from Microchip. Steve, Eric, welcome. Thank you for participating in our conference. What I was hoping to do is, Steve, if you could just start off with a few remarks. Microchip is very well followed, very well understood, but still, just a quick investment case for Microchip, and then we could go into the Q&A.

Sure. So I think we have performed quite well under very, very difficult circumstances. If you compare Microchip's performance in the last 2 quarters, March and June, and look at our guidance for September and compare all those 3 quarters to our competitors versus a year ago performance, competitors in microcontroller, competitors in analog and in FPGA, you'll find that we have double-digit lead over many of the other competitors when you compare the June quarter over the June quarter a year ago. That means we're continuing to gain market share in all of our focused strategic segments doing very well overall. And our gross and operating margins have held up phenomenally well in this environment. This is the highest gross and operating margins we have posted at the bottom of any cycle, and we are really a stone's throw away from our long-term target for growth as well as the operating margin. And later on, we can describe sort of a path to how we get there, which is really not too far away. So in terms of the end markets, auto market has been really weak in the June quarter and is recovering relatively strongly in the September quarter; while the data center market, which was very strong in the last year, is showing some signs of softening. The other markets are kind of no change. But overall, when you put it all together, we have a guidance for the current quarter, which is 0 to minus 8%. And we mentioned in our last earnings call that we were seeing very strong bookings. Bookings in July were strong, and bookings in August were continuing strong. So with another 3 weeks gone by here, August bookings have continued to be quite strong and probably strengthened further by some amount compared to July. So the backlog is continuing to fill. The point we will make is that the letter we wrote in early July could have not been better timed, it could not have been better written, and it could not have had a better effect. It really educated our customers that the lead times were not as showed as they thought, and they need to be giving us not only the shorter-term orders, which we needed to fill the quarter, but also medium- and longer-term orders so we can serve them better by building the product ahead of time. And they are doing so. So the bookings have been quite strong, and we've been getting medium- and longer-term bookings in addition to the short-term bookings we need to fill this quarter. So with that, I'll pass it back to you.

Okay. That's a great start. And Steve, this is a context that you brought up, which is the year-over-year performance versus your peers. And I think on the earnings call, there was some confusion from our side, buy side as well as sell side, in the commentary that you had given about the backlog and the near terms and implications for visibility. So if you put that in context with the year-over-year performance and look at the guide that you gave versus guide from some of the other peers who had -- year-over-year were down double-digit over the last 2 to 3 quarters. So if you could just kind of clarify some of the confusion. And what I mean by that is the comparison was made on backlog entering this quarter versus the backlog that you had in the last quarter and then the short-term orders and then how that translated into visibility for Microchip heading into this quarter. So if you could just throw some light on those comments, it would be great.

So when you look at the year-over-year performance, and especially the performance last quarter, some of the difference is related to the end-market mix, and the rest of the difference is related to our stronger performance, better relationships with the distributor, our total system solution methodology that's working very well and gaining share. So it was a combination of it. Last quarter, the automotive market was very weak. So some of the competitors that have a higher percentage of the business in automotive did worse. And Microchip also had a significant content in data centers, which was very strong, the data center contact -- content that some of our competitors did not have. So when you switch to the September quarter, there was a little bit of a reversal. Auto market is bouncing fairly strongly, so they would bounce stronger from a weaker start. Last quarter, they were worse. And in Microchip's case, we're benefiting from the automotive market, but the data center market is rolling over. So those 2 are kind of balancing. And I would say the third factor is some of our competitors were willing to make a call that COVID-19 is behind us, and the bottom is behind us, and it's all wonderful going forward, a call we were not willing to make. It doesn't mean we're doing worse. Our year-over-year performance is very good and will continue to be good. We're gaining share. But I do not know really where they draw their data that the COVID-19 risk is not there anymore. I mean it could resurge any time, and there have been some resurgences in some countries and some states. So it's just a question of what our comfort was with the unknown enemy versus theirs for us. It doesn't mean that it was affecting us more, it doesn't mean we were doing worse or anything like that. It was just a judgment call that -- when I make calls, I've always been correct, and I was not comfortable with making that call.

Great. Great. Okay. That's helpful. I'm just going to stay with near term very quick here. Huawei, it's a very small part of your business, does the latest ruling has any impact for Microchip?

Let me give that to Eric. I think he has been on the top of that. Go ahead, Eric.

Okay. So Huawei is not a material piece of our business. It's a 1% to 2% customer. So obviously, an important customer, but we have no customer that's more than about 3% of our revenue. We're continuing to evaluate the new guidelines with our compliance team, with our operations group and our legal team to make sure we're compliant. And I think, clearly, there's going to be an impact from it, but we are fortunate that it's not a large portion of our business as it is with some of our competitors. So we're continuing to manage through it. But at this point, it -- because of the size of the customer, it can't be a material impact on our overall operations.

Okay. Makes sense. And Eric, while I have you, maybe we can go into the recent transaction that was announced late last week. So if you could please highlight the major points from that transaction. And then how should we think about the impact of the balance sheet and the P&L?

Okay. So we were in the market last week repurchasing some of our convertible debt that was outstanding. We ended up repurchasing a principal amount of $414.3 million of our 2025 maturity convertibles and then $381.8 million of our 2027 convertible, so in total, $796 million. This was not a new financing. We drew down on our line of credit to repurchase the convertibles. And we've been saying now for quite some time, and starting back in March when we did our first repurchase transaction, that our convertibles are our most expensive form of capital in the structure. They obviously have very low cash coupons, but we're bullish on Microchip long term, and these are maturing in 5 years and 7 years out in time. And as Microchip performs well over that timeframe, and hopefully, there's stock price appreciation, these converts become much more expensive because of the equity dilution that goes along with them. So we thought it was prudent. Over the course of March transaction, what we did in May and June and then what we've done now in August, we've repurchased about $2.4 billion of our convertible debt, and we think that is very positive activity for the long-term dilutive effects of these converts that they have on earnings per share.

Okay. And back to the operating model and how much cash you generate. You've been aggressively delevering the model, your balance sheet. So just remind us where we are in the leverage ratio. What's the target? And then what's the right way to think about when do you get back to share repurchase?

Okay. So over the last 8 quarters, we have paid down about $2.6 billion of our debt. And we acquired a bunch of debt when we did the Microsemi transaction. And so we're making good progress. Obviously, EBITDA has been a bit challenged for us and the rest of the industry with the revenue headwinds that we face with the U.S.-China trade war and now the COVID pandemic, but we are making progress. We ended last quarter at net leverage of 4.24 and really have a target to get that at 3x or less over the coming quarters, and it will take us some time. But really, all of our cash generation beyond what we're paying to shareholders in dividends is being used to reduce debt. Last quarter alone, we took out $394 million of debt. This quarter, there'll probably be another $300 million of paydown, and we'll continue with that until we get down to roughly 3x leverage. And that's not set in stone, but we're very focused on deleveraging, and we think that there's some multiple expansion that will happen as we delever and take that risk off of the balance sheet, and investors will appreciate that.

Yes. And so the right way to think longer term, what's the right way to think about the debt level or the capital structure of the company?

So yes, you should really view Microchip as progressing towards being a stand-alone investment-grade company. And so getting to this 3x levered or below is where you should expect us to get over the course of time. And we can talk about M&A later if you want, but right now, we're really focused on continuing to integrate Microsemi and get the debt levels down. And I'll pass it to Steve to see if he wants to add anything else to that.

Well, certainly, we could talk about the M&A now. I think we've been very clear with the investors and analysts that our #1 priority is to complete the digestion of Microsemi and bring our leverage down to reasonable level, really, below 3, before we will do any other sizable M&A. But the second is, even when we get there, it isn't clear to us there is another large acquisition because industry is, number one, largely consolidated. And there are some players left, but the way the valuations are, especially last couple of deals, Cypress going for 10x', and similarly, Maxim. Those are not the valuations where Microchip will find them appealing, and we'll never pay that kind of price. So we are completely geared up to be able to grow organically with all the assets we have accumulated using a total system solution's approach and focusing on the 6 large megatrends that we have defined. We think there's a substantial opportunity to grow organically. And if an acquisition comes, it has to be at a reasonable price. And if the valuation stay where they are, then we're happy growing organically.

Yes, it's -- that was going to be my next question. So Steve, thanks for addressing that. And given the cash-generating capability of the model, and if I look back at -- I think it was Microchip 2.0, where you added the TSS part, which obviously you had been working on before. M&A was a big part of the strategy to add as one of the pieces is to add shareholder value. So if that falls lower down on the priority, then should we then expect more aggressive buyback post the 3 -- less than 3 leverage and a higher return via dividend?

Well, I mean the answer to that is yes. We will be generating a very sizable amount of free cash flow as we get down to those leverage numbers. And if we do not identify M&A -- like I said, there may not be an M&A -- then that cash will go in 3 different places. A portion of it, you could further pay down the debt. You could do some buybacks. And you could more aggressively increase the dividend. Those are not the discussions that need to happen today. It's a one-meeting discussion with the Board. As you get there, then you can deploy those alternative means of giving cash back to the shareholders.

Yes. And the next question very related to this is the operating model. When I look at what your long-term operating model is on -- both on the gross and the operating margin side, you're really within shouting distance of that. Now we're not expecting you to lay out a new long-term model at our conference. But how should we think about the levers for both gross margin and operating margin? Obviously, revenues help a lot. But are there other initiatives on whether it's on the -- streamlining the supply chain. And I think there is some manufacturing coming in-house, which should help as well. So just kind of help us understand, longer term, qualitatively, how should we think about the business?

So let me have Eric take that one.

Okay. So our long-term model is 63% gross margin and 40.5% operating margins. And this last completed quarter, we did 61.7% gross margins and 38.6% operating margins. And for the current quarter forecast, we're guiding September to about 38% operating margins. With the reduction in revenue, gross margins are expected to be about in the same range as last quarter, but OpEx as a percentage of sales will go up a little bit, but dollars are coming down. So -- let's start with gross margin. So in that 61.7% deliverable on gross margin last quarter, we had almost a $14 million charge for underutilization charges. So what we need, do we need to grow back into our capacity. And as we do that, those will slowly fall away. So that would add close to 1 percentage point, a little more than 1 percentage point to gross margin. And then on top of that, we're continuing to make investments in bringing some more assembly and test activities in-house that are currently subcontracted. Those are small incremental improvements that we gradually do in the business. So you can consider that just ongoing activity. And then we've got some other projects involved in terms of repurposing our Colorado factory that's on 6-inch wafers and making that a more specialized factory and moving some of the high-volume production that's done on 6-inch, moving that to our more efficient 8-inch factories in Arizona and Oregon. And so we kind of outlined back in November of 2019 or November of last year that this is a pretty large opportunity for us. So as we work through these things, we have high confidence that we can get to that 63% target. And we haven't got there yet. So again, not updating the model, but we feel very good about that. On the OpEx side, you know that we've taken some short-term actions. We've got our employees on a pay cut today. We implemented that back in April in what Steve described as anticipating a Category 6 storm coming. And maybe that storm has been downgraded today. Maybe it's Category 4, right? We're all experiencing the effects of the COVID pandemic. But I don't think it's hit as hard as we were fearing at one point in time. And so we've committed to our employees that, that pay cut will go away come January 1. You saw that our OpEx back in the March quarter was about $336 million on a non-GAAP basis. And in the current quarter, I think at the midpoint, we're guiding to about $298 million. Some of that is the pay cuts. Our bonuses are reduced currently. We've made reductions in other expenses. Just everybody has really rolled up their sleeves and cut expenses to the extent that we can while still making the appropriate investments in the business to drive the long-term health from a product and customer support activity. Some costs have come out of the system, which I think are more permanent in nature. I think a lot of the travel costs probably won't come back to the same level that we had them before, and we will continue to operate the business appropriately as revenue starts to climb and make investments where we think they're appropriate to again long -- drive the long-term health of the business. So some people call the long-term model conservative, but we haven't got there yet. And there are some expenses that need to come back into the model over the course of time as these pay cuts sort of essentially return to full level for the employee base.

Yes. And I think, Eric and Steve, you guys were very aggressive very early on in addressing the OpEx. So when the costs come back, what's the right way to think of how much of -- what percentage is permanent that will come out of the model versus, as you said, a lot of these costs will come back starting next year?

So I think essentially, all of the salary cuts and royalties return. They come back. And some portion of travel doesn't come back because some of the efficiencies we have gained through tools like these conferences and others are probably here to stay. But some of the expenses will come back. I don't think we have dollarized all of those. But salary cuts will come back. Cash bonuses will start to be given again. Some efficiencies in operating expenses, we will keep, and the others will come back.

And this is more a longer-term question, but it relates to what we are going through, I think, with 6-plus months of the pandemic. Steve, how has it changed? How you view a customer interaction, running the business, supply chain, design activity? What -- from your viewpoint, how are you adjusting to some of the changes that could last with us for a long time?

So I think our employees rose to the challenge a lot better than anybody could have imagined. I worked out of home now for 5 months and so have many, many other employees. And initial thought was, my God, face-to-face communication and personal touch is needed. And I would say it is still needed longer term, but years of camaraderie and leadership and working together and teamwork and knowing each other or one another and really a long-term tenured employees and executive teams that have been together for a very long time provided for really the avenue to be able to not see each other face-to-face but through these video calls and others, keep the momentum and keep the business going. And same thing in the engineering teams and design teams and all that. So only workers that are going to work are largely production workers and people who have to work in the labs for validation, characterization and all that. Majority of the other workers are working out of home, and they adopted working-from-home so seamlessly that it has been actually quite surprising. And I think part of the reason for that is really long-tenured employees and years of working together and trust and all that, that we have gained over the years. Now in terms of customer interface, our customers have told us that Microchip provided the best support during these times. And I don't know what the issues are, how they're comparing what others are doing and not doing. But if you look at -- one of our field apps engineer would serve maybe 1 to 2 customers a day. Many of them are in large cities and driving in Seoul or Tokyo or Shanghai, all that and would take 2, 3 hours to get to a customer, and then you have a meeting, and then it's too late for the next meeting. So sometimes, it's only one. Other times, it could be 2. Today, they have turned all that commute time into being with the customer on tube like we are, on a second tube watching a circuit board or a software and debugging it online with a customer. And they're serving 5, 6, 7 customers a day, having meetings every hour or so like we're having today. And there is a level of efficiency we have never thought about in customer support. And I think some of that will stay. So our customers have applauded us regarding how well we are serving them, and that has resulted into stronger design win, larger funnel, more TSS opportunities, larger success in TSS. And I think when all that turns to revenue, we'll see some dividends being paid.

Okay. One -- switching topics and relating to gross margin, and your inventory management approach versus that of the peers is very different. You have chosen not to build inventory. And I think last earnings, you were right bang in the middle of your stated goal, I think, 117 days. The bigger analog companies have all built inventory, but there's not that many of them. Now you guys are microcontroller versus analog, but you have analog as well. So -- and plus your channel inventory has been very lean as well. So just kind of help us understand your approach in managing inventory, both within the channel as well as on your own balance sheet. And how would you help us contrast with what those other diversified guys are doing?

So our feeling is that when you have high inventory -- and we have had high inventory prior cycles, in 2009 and some other cycles -- when the industry recovers, you have a very significant headwind because you're sitting on a large amount of inventory and you're not ramping the factories to meet the increase in demand because the inventory is then coming down with the good times as the times come. Microchip, unlike any other cycle, is sitting on an inventory at target at the bottom of the cycle. That has not happened before in any other cycles. And part of the reason is that we were in the middle of a fair amount of restructuring, trying to restructure our Colorado fab from a high-volume fab to a boutique fab and transferring that stuff to our other larger fabs. Also more than ever before, we were bringing a bunch of stuff inside in assembly and test as well as in fab, which has consumed the capacity in our factories, which ordinarily would not be consumed, and we would be building inventory with it. Or we will be going more idle and have larger underutilization charges. So this time, with that restructuring and with that capacity being used by bringing stuff from outside to inside, we have achieved 2 things: number one, we have right-in-the-middle target inventory. And number two, we have very high gross margins at the bottom of the cycle. So having those 2 things as the stronger times come and we are able to ramp our fabs and assembly and tests and remove the remaining utilization and then go above that, I think you will see a level of wind on the deck on gross and operating margins, where our competitors will see headwind because they're sitting on a really high amount of headwind.

Interesting. And this also speaks to the structural change, a, the industry has seen, and b, that you guys have done. And I think I asked the question a couple of earnings calls ago when I look back at '08, '09 and where the gross margins were versus where you guys are sitting at right now. The revenue mix and the product mix is a lot different. But can you just help us understand structurally? And I'll bring in microcontrollers where, for the longest time, there was a bear case that Microchip is going to lose share. But 8-bit has continued to have a much longer life than what folks were predicting back a few years ago. So, a, structurally, what have you done to get to this kind of cost structure where we are at the bottom of the cycle? I think it's you as well as the other big diversified guys have also demonstrated that. And then can you help us understand in -- at least in the big core parts of Microchip, what is happening? And this is not just a one-quarter thing because share shift doesn't -- don't happen in a quarter. But if I look back at the last several quarters, the MCU business has done -- I haven't looked at the last data, but the MCU business has done better than peers.

So I think our strategy in microcontroller has not been 32-bit. Instead of 8-bit, our strategy has been 32-bit in addition to 8-bit. Others made the call to stop investing in 8-bit entirely. And every time there is a socket, go after it with a 32-bit microcontroller, we have found that there are lower-end designs today that will continue to use 8-bit forever because they don't need to be connected to Internet, they don't need additional software, they have a very small footprint, they are simple socket where you're really controlling something very simple. And trying to always throw a 32-bit for every application is really not the most efficient. And we have found that we were right, and we have essentially huge share in all those emerging sockets where our 8-bit business continues to do well. I think we have said that a few times, but it doesn't seem to stick because somehow complacent investors are fixated on like we only have 8-bit, like we don't have 32-bit. Our 32-bit business is larger than 8-bit. Did you know that? We have said that multiple times. Did you know that, Ambrish?

I actually did not quite catch on to that, Steve, but I am also not one of...

So I made my -- so there I made my case. We have said multiple times that our 32-bit business is now larger than 8-bit business. Analysts haven't caught on to it, and investors haven't caught on to it, and therefore, the risk of the 8-bit kind of continues in the background as a bear case. We have deployed a strategy of 32-bit in addition to 8-bit. We have very large share in the 8-bit business, and 8-bit is still growing. In 2018, we did record on 8-bit. And it has been slow last 1.5 years because of the combined effects of trade war and now COVID-19. But we have a very good 8-bit business, and we are confident that it will grow again as we come out of these COVID-19 issues. Meanwhile, our 32-bit business has grown very large and is now larger than 8-bit business.

And we have a very strong 16-bit business also. So we've got low end of 8-bit, the high end of 32-bit, a very, very broad portfolio.

So help us create a situation where 90 days from now, there's no analyst or investor that doesn't know that our 32-bit business is larger than 8-bit.

I will make sure that I will write it in my note, but -- and I'm not defending myself. My investment case for Microchip has never been about 8 and 32 and 16. It's a little bit broader than that. So -- but yes, we'll make sure that it shows up in our note. And now it's registered, Steve, it's going to stay with me.

Hopefully the investors who are listening to this -- we can't see them. I think they're all dialed in or something. Whoever is listening to it, make sure it registers with you that our 32-bit business is larger than 8-bit and doing extremely well.

It is registered and imprinted. And we'll put an imprint as well. Why don't we stay with the -- actually, you didn't address the structural part of the question, the structural changes that you have made within the company to get to where we are at the bottom of the cycle where we are sitting at these kind of operating margins and compared to 10 years ago. So if you could just touch on that as well.

Talk about the structural changes, is that the question?

Yes. Yes.

Yes, do you want me to start, Steve?

Yes, go ahead. Yes, go ahead.

If you look back in the 2008, 2009 downturn, we were much more dependent on internal manufacturing. And internal manufacturing is a strength of ours. But when you have a downturn, if 90% of your production is done in-house, your underutilization charges are going to be higher than if only 40% of your production is done in-house, and that's where we are from a wafer fab perspective today. So that -- I think that's a pretty big structural difference. But it's not like our capacity has gone down. If you look 10 years ago, we were roughly $1 billion in revenue. And today, we're $5 billion plus. So the company has just changed dramatically over that time period. So that's one thing. What would you add, Steve?

Well, I would say the same thing on assembly and test, where 10 years ago, 95% of our test was done in-house; 70%, 75% of assembly was in-house. Today, those numbers are -- Eric, if you can give the numbers, Eric?

Yes, 55% for test and about 45% for assembly.

So there, again, it gives us the ability to keep our factories full and bring more stuff from outside to inside to keep it full and really having lower incremental costs. So those things have really done very well to maintain the gross margin. In addition to that, we've also acquired companies and then did things in those businesses. When we bought Atmel, it had a low gross margin, was at breakeven operating performance. Today, the operating performance is in high 30s. So those things have worked very well. The ASPs have held up very well. I think today, the customer does not get a year-over-year ASP drop every year like it was true in semiconductor industry for 3 or 4 decades, and customers are now being trained essentially that here is the price of the product. And next year, you get the same price and not lower. We used to only succeed about 10% of the time 5, 6 years ago. Today, we succeed 90% of the time. So ASPs have held well and costs have continued to come down through various ways, and we have been able to show that on the operating margin. And finally, we have taken some idle or inefficient capacity off the line. We closed down a small fab in Bend, Oregon that Microsemi had. We are in the process of closing down a small fab in Santa Clara. We closed down the old Micrel fab, a 6-inch fab, in San Jose. We also closed down a small Supertex fab in San Jose. Those were the changes that we made in the last 5 years. And we have restructured Atmel's Colorado fab. We're very -- it was a very large 6-inch fab and inefficient with low demand, and we have restructured it to essentially be a boutique fab producing discrete products and analog products and a bunch of other stuff and transferred the high-volume manufacturing to our 8-inch fab, which are much more efficient. And we described that will add about $65 million in accretion over the 3 years or so, and we are well on our way to really achieving that. So those are many of the things we did. And some of those things, we started when times were still good, and it has helped us. As the times have been bad now, some of those things have come to fruition, and it's showing up in our wonderful gross and operating margins.

Right. Yes, clearly a lot. I'm just going to pause here and see if there are any other questions. That microcontroller question was actually from the line. And I'll follow up with Eric later on to get the actual numbers, break-up between 32, 16 and 8. But while I wait for questions, do you want...

You won't even get it when we are breaking it out. So don't try to ask him.

Okay. And you won't see it...

Somebody breaks it out. We're not going to break it out. I think you could possibly get some number from some outside sources, but we're not confirming and not breaking out the numbers.

Okay. And you definitely won't see a title from my side saying Microchip, the 32-bit story, because I think you're broader than that. So a couple of questions. And this also comes back to what was on my mind as well, is that TSS approach. What are some of the secular drivers that you are focused on with that approach where you're bringing in -- from microcontroller, you have expanded well beyond that. You have had analog, but there are other pieces that you bring now on the board. So just remind us what the TSS approach is, and what are some of the areas that you're focused on?

So stand-alone also -- but also when we acquire a company. So take the example like we bought Microsemi 2 years ago. And if you look at a Microsemi FPGA reference design, when we first bought them, they had 14 analog chips around it providing power management and other with the central chip being a FPGA chip from Microsemi. And out of those 14 chips, 2 of them happen to be Microchip. They were not preferring Microchip. They were an independent company. But the other 12 were chips from other analog companies that you guys name every day. Fast forward, today's reference design, 100% of those chips are from my Microchip. So it's a question of, number one, having a broad enough analog line to have those chips available to design with; and number two, having a focus where the entire company, every business unit really signed up to that charity begins at home. You are going to design Microchip products, I don't care what the preference of the FPGA engineer is, whoever is doing that reference design is going to put the Microchip part first. And that is happening. Can you put that TSS slide, Eric?

I will try to do it.

So while -- let's see if he can do that. So using that, essentially, every time we have an operational review of any business unit and as they are talking about their design wins, one of the things they're supposed to cover is, along with winning design from their part, whether it's FPGA, whether it's 32-bit microcontroller, what other TSS content they brought in? If they don't cover that, they get beaten up. So here is a current server design with a major server manufacturer that you will know very well. And there are 5 or 6 boards here. There's a front panel. There's a display. There's a motherboard, PCIe card, NIC card and backplanes. And years ago, we would have 1 or 2 microcontrollers on this board. Today, it's populated with our chips. So there's MCU8 in it. There's 2 chips of MCU8, 3 chips of MCU32, WSG's wireless system group, UNG's networking group. There are 2 USB networking chips, TCG's timing. There are 5 chips from there. APID's analog power. There are 2 chips from there. And CPG's computing product group, and there is one 32-bit chip from a computing product group. And then when you look at the next design, which is not in production, you'll just see how much more populated it is. A few more business units have content in it. Each business unit have a higher content in it. I'll have Eric go back and forth a couple of times so you could see the contrast, now and the next design. So this is sort of what TSS is all about: one, we today have the product breadth available designed internally plus acquired through acquisitions to be able to achieve this; and number two, have a structure in the company and a mentality of the engineers and training provided so that they design all these boards with our chips and then taking it to the customer, training them, in many cases giving them a ready-made board that some customers wouldn't mind just copying it.

It's -- this is a great example. Just a question on one of the acquisitions, one of the assets you bought through Microsemi is FPGAs. Is FPGA -- and there's obviously a lot that goes along with it, and it gave you access to the aerospace and defense industry. How do you view FPGA? Do you need to build on that organically? Or does it just fit in complementary with what you're doing and you can bring in more Microchip products into that end market?

No, so FPGA is about closing on $400 million business, not quite there, although we've had a quarter over $100 million, I think, or close to it. So it's really sort of the third-largest FPGA line behind Xilinx and Altera. Although, we don't see a lot of competition because we largely occupy the mid-end space. And we compete with Xilinx, Altera at the boundary of mid- and high-end, and we compete with Lattice at the boundary of low-end and mid-end. But the space we occupied is largely very, very protected. It makes very high margins, very high gross margins, very high operating margins. Yes, a good portion of the business is in aerospace and defense, which also drives margins, but a good portion of the business is in industrial and communication and other sectors also, which also has very good margin. So we are building -- currently, the production is in G3 and G4, Gen 3 and Gen 4. And recently, in the last couple of years, we have introduced fifth-generation parts, and the sixth generation are under design. With each generation, we continue to move up. So with the next-generation, yes, we'll move up and occupy even more space and push up, competing some at the low end of where Xilinx and Altera might be. And similarly, we are expanding on the low end and pushing up against where Lattice is and take some of that space. So it's really elbowing out on both sides, push up and push down while keep the command at the middle-end space with very high gross and operating margins. I think it's been -- it's a phenomenal business, and it continues to do very well. And secondly, it's pulling in a lot of TSS. See, we could do what Altera can't do and what Xilinx can't do. They provide the FPGA, but they don't have the analog power, they don't have the connectivity chips, they don't have all the other stuff. So just like Microsemi, where 14 other chips were made by somebody else, today, those 14 chips are made by us. So the chip -- getting total system solution to the customer, not only selling the FPGA, but the entire power management and connectivity, the entire board, we are uniquely positioned to provide the total system solution above anybody else in FPGA.

I think maybe another thing I would add to that is Microchip also brings some new strengths from an end market perspective that Microsemi didn't have historically. And so we can leverage those strength, whether it's in automotive or another end market, and expand the places where Microsemi has a possibility of being successful with FPGAs and other products.

Great. Great. Listen, this has been fantastic. We are coming to the end of our allotted time. I'm glad, Steve, I got you fired up. And I learned something new in the process, too. But thanks again for participating. And if you had any closing remarks for us to walk away with, that would be great.

So my closing remarks would be that you can remember we are uniquely positioned to provide the total system solution to our customer because there's really no other competitor. Maybe TI comes the closest, but their microcontroller portfolio is becoming quite weak, as they're really not focusing on it for a number of years now. There's really no other competitor that has that breadth of microcontrollers, analog, connectivity products, FPGA, power management, memory and all that. So we are uniquely positioned to bring you the total system solution to our customer.

Great. And plenty of leverage to go into the model when things recover. Awesome. Thank you very much, Steve. Thank you, Eric. Take care, guys. And we'll be back -- for the rest of the listeners, we'll be back after a short pause. Thanks.

Thanks. Buh-bye.

Thank you. Bye.

Bye.