Microchip Technology Incorporated (MCHP) Earnings Call Transcript & Summary

Good morning, everyone. For those who need it, there's still coffee in the hallway. Thanks. I'm Chris Danely, your friendly neighborhood semiconductor analyst here at Citigroup. Next up, it's our pleasure to have one of our topics. Microchip, we have Ganesh Moorthy, the President, CEO. We also have Sajid Daudi, VP of IR.

I'll start with the most obvious question. One of the reasons why we like Microchip so much is it can't get any worse, right, with sales down the most among the analog games, and it's a very high-quality company. We don't think that the business model has really changed. Ganesh, maybe in your own words, why do you think that your sales have gone down more than peers?

So before I the answers, let me give you the obligatory -- during the course of this discussion, we'll be making projections and other forward-looking statements regarding future events of financial performance of the company. And we wish to caution you that these statements are predictions and actual events or results may differ materially and we refer you to our filings with the SEC for any risk factors that might impact our business and results of operations. The magnitude of the change has to be seen on both sides of the cycle. So we probably had some of the largest growth as well in some of the prior quarters. And all of that was a psychology of fear in terms of where customers were with their shortages. And we had this COVID-driven oscillation that began in '21 time frame -- 2021 time frame, drove a lot of people who were trying to get capacity, get supply for themselves. And did it overdrive along the way? Probably did. Part of it was customers were afraid. We had higher exposure to industrial and automotive. Many of those customers build end products that are quite expensive, didn't want to be held back by products that were -- the semiconductor components where they're much smaller in the [indiscernible] for them. And then the second part of what happened is their own business began to soften against their view of their business back in '21 and '22 and '23 when they thought the growth was going to be continuing at a rate, right? The overall macro slowed down. Interest rates went up. And so their view of their business began to slow down, their ability to keep the inventory became harder in a high interest rate environment. And so we're going through an unwind, first of the inventory and second of the macro weakness that they are trying to work on [indiscernible]. And in an environment where supply is plentiful, lead times are short, customers have a comfort level of not carrying inventory, not providing much visibility because they want to contain their own risk, which is driven by the macro at this point to some degree of uncertainty.

Just to dig in on some of that. I know you have 2 main businesses, microcontrollers and analog. Have you seen more weakness in either? And if so, why has one been weaker than the other you think?

In the aggregate, I don't think they're materially different. These products go together into the end systems that we are working with customers to get designed into. Another way, more of one content, less of another content. But really, we're selling them as total system solutions when we approach a customer. And I don't expect there's a material difference between the analog product. And those are just product line designations, less to do with end markets and applications.

And when things do bounce back, would you expect them both to bounce back roughly at the same time, any one more than the other? Any insights there from what you've seen so far this year?

I would expect when they come back, again, they come back together. The timing of any one customer or any one end market may be slightly different. But we sell as complete solutions into these customers' applications and markets, and they will go down together, they will come back together.

Great. And then let's just dig in on the inventory trends. Where do you think inventory is in the best shape out there? And where are you more concerned?

Customers don't report inventory to us. They maintain that as their internal thing. Different customers have different views of how strategic are they thinking about inventory. And many of the CEO conversations today, customers are as concerned about is there going to be a shortage in 25 or 26 and the next subpart of the cycle is where they're concerned. Not everybody is thinking along those lines. So what inventory a customer carries is a reflection of what business they're in, how strategic is their thinking process and then how much are they cash constrained in running their business and where they want to go run things at. But they're all over the place and where they're at. And it's not just customers, right? Some often, they have inventory too with their channels with their end markets and where they're at. So there's multiple levels of fish that's there. But the water level is declining every quarter as we undership the demand. And at some point, it will flip. And we're not trying to predict exactly when it will flip. But I think we're closer to the later innings of this inventory correction.

And why do you think we're closer to the later innings? What gives you some confidence there?

We're touching by time and what has happened over multiple quarters. And some of the patterns we see of people starting to provide more accelerated orders coming in with expedite requests, people pulling in orders from, say, next quarter or back in the next quarter. All of that would indicate that their inventory is getting to a point where by, they don't need to give visibility, they need to buy it to just keep building what they have in the short term.

Given that inventories are going down, do you think there's any risk that we see some sort of extension and lead times as things do start to bounce? I mean you mentioned a few pull-ins. Has that impacted lead times at all? Or is it now we're still in inventory like burn mode? how do you look at it operationally, I guess?

So for the period of time where at least in the early phases of any recovery, lead time won't be immediately starting to push out. Number one, we have inventory and we have inventory positioned in such a way that we can respond quickly. That's how we get positioned for turns, orders in an uncertain environment where we're at. And we have capacity that we have underutilization, for which we can turn on and add more. It all as a function of do we have enough of a lead time that customers start to place orders in. So what happened in '21 was long many customers simultaneously hit the accelerator. And so it takes time where production cycle times are closer to 5, 6 months, whereas we have lead times that are 4 to 8 weeks, and we maintain that difference by having strategic inventory position either in finished goods or to a greater extent in [indiscernible], which is finished with a wafer fabrication. And as long as the demand comes back at a rate that is at a reasonable rate, we think lead times will stay short. At some point, they could get tighter. But if I go pre-COVID, we ran at less than 8 weeks for years as a standard model in which we run. So I think that's the steady state we want to run it. I also believe short lead times are good not only for the customers, but for us as well, so demand signal is a lot cleaner because customers are not trying to predict where is the business going to be in 6 months or 9 months or whatever time frame things are going to be. They are able to adjust to market changes that are taking place, and we're able to adjust that as well. So in general, short lead times help everybody in the chain be adaptive to where the market is.

And you mentioned as far as what gives you a little bit of confidence that we're getting close to the bottom. You've seen some more accelerated orders and some pull-ins. Is that any different than what you had, say, 3 months ago or 6 months ago? Or doesn't this happen just to play devil's advocate? Doesn't that happen every quarter? About a year ago, but...

No, there is some. But the rate at which we see -- the number of expedites that are coming in, the number of pull-ins that are coming in, right, continue to grow -- and so that is a sign that more and more customers and even a customer not monolithic, right? They may have certain parts of the business on which they have inventory, but other parts of the business that is out of inventory and where it's at. But that rate of number of customers, number of times we're being asked to expedite is growing. And I think that is a sign that inventory is getting down to the bottom in many cases.

Is there any consistency or rhyme or reason where you're seeing it? Is it more prevalent in any particular geography, end market? Anything else, any sort of pattern you can see out there? [indiscernible] curiosity.

Harder to discern some of that because of how broad-based we are. I think geographically, as we said at the conference call a month ago, Europe is the weakest area. Europe is also in its -- not only has it got the weak macro, if you saw the German GDP numbers where it sat within the last week or 2, it reflects that. But also, they are in their holiday quarter, in the September quarter, so there's less production in terms of where that is taking place. But Europe is, it started later, but it's got the most headwinds in it. The U.S. has headwinds. I think the PMI numbers in the U.S. reflect that. I think it's 21 or 22 months. We've had a negative PMI 5 months in a row, I think, at this point in time in terms of the announcement of Street. There is stabilization we're seeing in China and the rest of it is [indiscernible].

A So just to get the geos down, you say Europe is weakest, even though it's a seasonally slow quarter for them?

Yes, I think it is more than just seasonal in terms of European weakness.

And would you say the U.S. is a little weaker than seasonal as well? And then China is the strongest? Has that been true for the last would you say a couple of quarters? Or any changes you've seen over the last 6 months out of any of those 3 geos?

I think the U.S. is -- again, seasonal at this point is hard to tell. We haven't had a seasonal quarter in like 5 years.

Up or down?

What's that?

Up or down?

Yes. So I mean between the different shocks in the system, there was tariffs that was covered, there's inventory correction. There really hasn't been something you could say, "All right, we're kind of in a steady state and you could call it." So I'm not seasonal for Europe only in that they have more holidays in the September quarter, and that's why I went there. Outside of that, I think there's weakness in the U.S. that I don't -- I wouldn't attribute to seasonality, but I would just attribute to industrial base is weak. And automotive is not as strong as it used to be at one time. And with respect to China, I would say it's more within the last 3 months, 3, 4 months.

Okay. And can you just remind us what your rough splits are between those 3 regions?

Yes, it's about, I don't know, 20 -- mid-20s in each of Europe and the Americas. The balance is in Asia and China, including Taiwan is probably close to 30%. -- Asia, excluding the China and Taiwan is a balance 20% or so plus or minus a couple of percent.

Sure. And then maybe switching gears to the end markets. Are you seeing -- you talked about the machinations of the geographies. Are you seeing any difference between, say, the auto or industrial or aerospace and defense end markets in terms of strain weakness?

So Aerospace and Defense has been a steady performer and commercial aviation is so strong. It's the smallest piece of A&D for us. Space tends to be lumpy depending on programs and where it's happening in defense, perhaps not always for the best of reasons, but it's strong in where it's at. Industrial, excluding Aerospace and Defense is weak. Automotive is still going through its correction. And it's -- I would say, from our perspective, is weak. The data center, on the other hand, is strengthening. And so not only for a while, we had the AI subset of data center, which is about 30% of our data center business, was doing well. And now the rest of data center is also strengthening for us. So that's a good sign in where it's at. And it's hard to see much improvement in most of the other ones at this time.

Sure. Would you say there's any difference between the auto and the industrial space as far as one weaker than the other?

Yes, industrial is weaker.

And why do you think that is? Do you think it's the underlying demand trends? Or do you think that as there's been a little more inventory build or both?

I think there are many causes, I think all of those are good reasons why, I think industrial also has large capital spending that is associated with it. As companies slow down capital expenses, et cetera, that slows our industrial down as well. Automotive tends to be more of a consumer spend for the most part. Of course, consumers are stretched for the higher interest rates and what we're taking pay. So I think there's [indiscernible]. But the industrial piece is certainly weaker.

Okay. And so you spend a lot of time with the customers, what are they saying as far as their own inventory goes? Do they want to take it back to normal levels, below normal levels? This is just like massive inventory every year not because we're already down 45%. Any insights there?

Again, as I started earlier, I think customers are very, very different in how they think about things. A lot of the end products that we end up in a high-value products, right, if you think of medical, industrial, automotive. I mean, these have semiconductor content that may be a couple of percent of what the total value of the end product is. And in that, you have a choice. So there are many customers that look, I don't ever want to be in the position I was in, in '21 and '22. And so I'm going to make the strategic investment of having inventory to ensure that doesn't -- in a semiconductor inventory. So there are many customers that I speak to who are still very much are thinking along those lines. Maybe they don't need to have excess inventory, but they certainly want to have the inventory that allows it to not get bottlenecked by semiconductors where they're at. There are others that say, "Hey, my cash flow is tight. My CFO is pushing on the what I could afford in inventory when interest rates were 1% to what I can afford when interest rates at 6% to 7% is quite different. And why do I need to carry inventory when supply is ample and lead times are short." So it all is a function of each company's thought process and what they want to go do. I don't think the need for fear-driven inventory is there. And so there was a trouble perhaps in the '21, '22 time frame where that fear caused people to try to find it's a bit of a hoarding that tries to take place. Now you couldn't supply everybody with everything. So it didn't get as big as it perhaps could have been. But I think people will be thoughtful about semiconductor inventory going into end products that are high-value.

And so what would be your overall take on just underlying demand trends in semis? When I say trends, I mean, do you think the demand has gotten worse over the last 3 months or stabilized or better or somewhere in between?

We're -- it's hard to read the teaming. So when you say demand -- I think you have to think about when we think of demand, we think of what's end consumption doing and to the best of our ability to read it because that's what ultimately drives the overall change that has to -- that helps to drain inventories. That's what drives the builds and all of that. And I think we're seeing, one, is inventory is draining. -- but the macro is weak. And people are uncertain as to what should they build for, what kind of growth rate should they assume? There was a time back in the '23-'24 time frame where many, many of our customers, "Hey, our demand trajectory has changed." We're going to be growing 10%, 15%, 20% a year -- and as much as we were skeptical on it, right, I mean, I have had multiple CEOs calling us, beating on us for products, saying, "You're limiting my 15%, 20% kind of growth rates." And we're going to project that for the next so many years. That's clearly gone in terms of where they are. There could be individual customers that are doing that, but not the entire market works out. So I think today, what is reflected through the chain back to us is in part the inventory reduction, which I think, as I said, is in the later innings, but in part and uncertainty of their own markets. And I think the uncertainty in the market is creating as much of an issue today where they don't quite know what should they build and to what extent should they do. And they will get clear on that. Hopefully, some of the uncertainties in the market, what interest rate is going to do, what is the election going to do, et cetera, will provide a degree of comfort and confidence on what should they plan for and workouts.

I have a sort of a macro question on that. I mean you were around the last time rates went down. There was a lot of speculation -- they're going to go down again. So do you think that rates going down is going to help the tech or semi industry in general? And do you remember like what happened last time the rates went down?

Yes. So I think the how the rates go down and how much they go on? Is it 3 cuts or 2 cuts or. I think those are less -- those are -- it's in the right direction. When rates going down is it helps build confidence on the direction of where things are going to be over a longer period of time. So we're less worded about, is it a September or October cut and is it 0.25 point or 0.5 point, et cetera. But if a market gets confidence and if customers get confidence that the trajectory we're on is going to create economic stimulus. That's the more important outcome from this.

That's interesting. Next question is a little difficult. I have to ask it, [indiscernible] Research here. We mentioned that your sales have fallen more than the competitors. I'm sure this keeps you up at night. Why would this not be share loss? Or do you think there is some share loss in there?

So we haven't lost the customers that were generating the revenue for when we were going through the higher quarters itself, right? We know and we're in touch with all of them. We have an idea of where things are at. So I don't think this is a share loss issue as much as it's a decrease in customers bring their inventories down and it's a slowdown in their businesses that is driving it. And we have to look at -- the way I think about it is if you look at the area under the curve, which is how much revenue was generated and you start from January 2020. So you used the December 2019 quarter as an index of normality, whatever that is, pre-COVID. And you look at the next 18, 19 quarters and you've [indiscernible] for any of the folks that are out there you'll find the area under the [indiscernible] by revenue. So some of you have done it earlier, some we have done it later. It's going to be almost identical for everybody indexed to the December quarter of 2019.

Okay. Would you say that your visibility now versus where you were, let's say, 3 months ago, has that changed? I mean, can it get any worse? Have you seen any signs that it's getting a little bit better? Or is it about the same? Is it still like just pure charcoal when you look at the order rates out there? What...

Visibility has not improved. And I think that's because the environment hasn't improved. What has improved is the level at which a number of expedite requests are coming in and et cetera. So I think that's the best way to think about it is I don't think there's any improvement in visibility at this point in time.

Yes. Can it get any worse, by the way.

Anything can get worse, who knows. It doesn't feel like it's going to -- that's where the issues are going to be.

Yes. Here's a question you probably haven't heard for a while but turns percentage. Maybe talk about what the typical percentage is and what you're looking at these days? And do you expect that to change at all?

Yes. So historical turns percentages are not as relevant today. I mean it's an important about data point. So if I go back to pre-COVID, we used to run 30%, 35% turns in a given quarter -- but I think turns in an environment in which there's inventory is different where you may think, I should be back at 30%, 35%. And -- but if there is inventory to drain, those are not going to show up as turns in the quarter. So we got to manage for all that. I think in our guidance is built in what we expect in terms of turns as well as what do we expect in terms of inventory that's draining. But we have to keep both those in mind as inventory drains. It would just be looking at historical alone.

Got it. So you mentioned that you've seen some signs that we're getting close to the bottom. I remember back at one of our dastardly competitors' conferences a few months ago. I think it was in May or June that you talked about things some green shoots or signs that things were -- could be bottoming. Are you seeing any difference now versus then? Because obviously, since then, your sales have fallen again, maybe compare and contrast, do you have more confidence now? Or is it just all right, we're seeing a few signs and who knows?

I think at the time we looked at it, the one difference was that bookings were starting to grow month after month after month. And we're starting to see highest bookings rate and even prior 12 months, multiple times. I think bookings have stabilized. And I think in part, it's because people don't need to book when they have an environment in which lead times are short. And often, bookings reflect lead times. You get tons of bookings when lead times are long, even if they're wrong because people really don't know what they want, but they said I better get in line and order something, which is why I think the short lead times are helpful to give people an ability to book in such a way that it reflects the best demand signal that they have on it. We have also given customers more leeway in terms of cancelable windows, we reduce all that stuff to give them more flexibility in where that is. So while the turns are still high, the bookings are more stable. And I think bookings are going to stay that way for until there's more confidence that customers have in the environment, which goes back to the interest rate and other uncertainties as they get more clear. We'll come back to that. And of course, as people had bottomed, more orders just need to be placed to get the consumption, right? The up cycle will play out in 2 phases. In the first phase, it will be people getting back to consumption, which is higher than where today's shipments have taken place. And the second phase, it will be around people's growth driven by macro factors adding to whatever consumption is that goes from there.

How are you guys managing inventory? Because we've started to see some of your peers bounce back. They're starting to see some above seasonal order trends, which you would expect to happen -- and how do you sort of manage your inventory, right? You don't want to take it down to nothing because then when the orders bounce back, your lead times out to Infinity. So how does that work as a CEO?

So it's got multiple facets. So we've got a part -- 60% of our business is done through the foundries. And so that's really not reflective in our factories. It's working with our foundry partners on how are we managing -- there are some challenges there because people are also end-of-life certain of our processes and we're building end-of-life inventory and high gross margin products, et cetera. It has the last count of like 19 days or something in our days of inventory that, that was carrying. There's our factories. Our factories are running underloaded. And every quarter, we have an underutilization charge, but we also have things that we write off because it doesn't meet our rules. And -- but it's all good inventory that we build -- and in every cycle in the past, that inventory has been something we monetized and get generated cash and working it and we'll do that again. So where we are today is we do have high inventory. It is at the best possible place for it to respond to a change in demand. It's not -- this is not very substantially different, although not at the same magnitude as where we were in 2020, right? 2019 was a weak year. We built inventory. And then as we went into 2020 post COVID, a lot of that inventory is what helped us to in the early phases, response to demand. And then that gives us some time to turn the factories back on and a thing. But keeping the factories running allows us to have equipment, people, et cetera, are ready to go, and they just adding more people to run more systems and improve the utilization of where it is to where we would like it to be.

So yes, I think -- I don't know if you guys are still doing furloughs, but maybe talk a little bit about utilization rates now. Are they pretty much as low as they're going to go, where your inventory days are where you expect them to go? Will you keep a little bit more, even though we're at the bottom of the cycle, just in case the order rates bounce back? How do you look at that?

So we did furloughs in March and June quarters. We are not doing one this quarter. We are at a place on inventory where we're comfortable with where it is high and -- but it's also the place I think we need to be positioned. So days of inventory is a backward-looking indicator and it divides by the lower revenue -- the lower cost of goods reflected by lower revenue. And so that number can change quite dramatically as revenue starts to grow because you get both consumption of the inventory, but also higher -- the revenue and COGS grow as well. So days of inventory is 1 indicator, but we also need to be thinking of forward looking, where do we expect revenue to be? And then how does that inventory support, not what we needed to go. So I think we're in a place we want to be with respect to that. In the back end, our packaging and testing facilities we use -- rather than furloughs, we use shutdown days. So we'll add a day or 2 here and there and through the course of the quarter, have multiple days that are there. That still is a part of what we do to manage because you don't want to wrap packages around the silicon or get them customized to what the final product is going to be until the demand signal is clear.

Because there's no furloughs this quarter, did you lower utilization rates or where utilization rates flat this quarter versus last quarter?

Utilization rates are roughly the same. We're not trying to meaningfully adjust them and it comes to where they are. Now we can still be running less product through the factories depending on how many wafers we start and what cycle times they're running through them, et cetera. But I think it will be roughly flat with last quarter.

Okay. And so what will that do to inventory days? Do you expect inventory or inventory days to go up, how much this quarter? And then...

So I don't have an exact number. But by definition, if our revenue goes down this quarter, inventory days will go up because of the way [indiscernible] in that's time.

Got it. Now given you have inventory, there's inventory in the channel, this has been going on for a while. Are you seeing any pricing pressure? What should we think of as far as price goes for Microchip?

Let me take a moment more on inventory just to give some comfort. So vast majority of our products, a very, very long-lived products. So these are products that have 10, 15, 20 years. And that's where we'll build inventories on things that have low to no risk of [indiscernible]. So I want you to take away that the high inventory doesn't have an [indiscernible] risk, it does have some timing over which you have to go draw down. And our pricing is not a -- unlike a more commodity market where there's high elasticity on pricing that adjust what happens. This is a much more inelastic market. It's much more driven by where design takes place and the length of time it takes. So we could reduce prices tomorrow and have no impact on revenue in the next 6, 12 months of time, and where it goes. So pricing is much more disciplined. Pricing when you're fighting for new designs has always been competitive. This is not something new. This is the way it's trending 30, 40 years of time. Everybody is using the best new product capabilities that you have because price is not the only decision-making criteria. It's cost of ownership. It's the technology you bring, you see the value that you bring in terms of where it goes. And we remain competitive at the point of design.

So are you seeing any more pricing pressure, say, now versus a year or 2 ago?

Every good purchasing manager is going to ask you for lower price. But that is not something everyone necessarily to respond to. If you're talking about at the point of design, of course, everybody is going to be price aggressive there. But it's not a new characteristic. With the exception of 2 years during COVID, when new designs got suppressed because customers were just playing triage, they're trying to get their production running, get their substitute products, et cetera designed in. So at the point of design, it remains competitive, has been competitive, always will be competitive.

Will that go away once whatever the upturn starts? Will take a couple of quarters based on your experience in previous cycles? Does it take 1 quarter, 2 quarter, 3 quarters of the upturn in better orders and sequential increases in revenue before the price structure goes away? Or do you expect it to just remain like fairly constant within this level as long as nothing crazy happens?

Again, to separate out at the point of design, we're always competitive, right? I mean you say pricing pressure, it's not only price design decisions are made for 10 other reasons and price is one of them, but it's not the only reason why. That is the way it always has been historically, it will be, and I expect it will be going forward as well. The only exception was during COVID when fewer designs were being done. With respect to pricing from a production standpoint, it's stable pricing, right? We work that into how the negotiations are at the point of design. And I'm not expecting that we -- this is not a commodity market where you raise prices just because it's supply constrained and you lower prices just because you have extra capacity to...

Great. I have 1 question. So you said that you had furloughs in the March and the June quarter, but none this quarter, right?

Correct.

Why was that decision made? Why no furloughs this quarter?

Well, it's the way we want to manage our business. We had gotten the head count down to where we needed to go. We've got the factories -- we need the factories to run at a certain level so that it builds and has consistency for the up cycle. So we've got to take these things not on a quarter at a time, but really on a longer-term basis where it needs to go. Sometimes in our shorter-term goals we're trying to achieve to go with it as well. So that's all built into what we needed to do.

Great. Let's sum a little more longer-term question, get of the short-term stuff. 2025, what are you most excited about for next year -- as far as longer-term trends go?

There's no single item that excites us or not. I think we're excited by the amount of innovation that semiconductors are driving in the markets that we play in, in the megatrends that we are focused on, in the way in which customers are using the entire capability of Microchip. So no matter what end market, what megatrend we're in. If I look at the design-in activity, I expect the outcome from that design-in activity as it goes to production and the fruition into our revenue growth to be something that is what we look forward to. Hard to see that in today's environment when revenue is constrained and where it's at. But that's what we look for is to create the future, build the pipelines, win the designs to drive growth.

I was hoping you're going to pull out a crystal ball and say we're looking forward to the upturn. As far as the gross margin drivers from here, were #1 is utilization rates. Talk about anything else going on? And then do you think that your gross margins can eventually get back to the previous peak?

So on the latter, absolutely. We think we get back to 68%, which is our target. How we get there will be a function of how the upturn plays itself out. But there are many factors that go into it. So underutilization, reversing that is certainly a key piece of how that goes. Product mix continues to be a way in which we have different gross margins in different product lines and then how that all changes in time will be there. And then just ongoing improvements and yield, improvements in factory capabilities and all that, which is kind of the nuts and bolts of how semiconductors run, all drive there. So there's no piece of the gross margin drivers and numbers say the fourth piece is pricing discipline. So maintaining that pricing discipline and as important as all the operational efficiencies on the where we get to. But gross margin is in our blood. It's been -- it's a mindset in what we do all the way from what we do operationally, how we design our products to how we price our products. And so I expect we will get back there as the revenue goes back towards where it was before.

Great. We have a few minutes left. I'd be remiss if I didn't see if the audience had any questions out there, going once, going twice, I'm fine, continuing to flat my gums and pester the [indiscernible] Ganesh, yes. Do we need a microphone or...

Ganesh, a quick one. On the GM side, one of the biggest -- one of the contributors was as you started doing more packaging for some of the acquisitions you did. Are you guys done there? Or do you have any more to go on that side? And then just a follow-up, you just mentioned that pricing discipline while there's a few statements ago, you said that it's pretty in the last. Is it pricing discipline for a certain part of product? Or is it like -- I mean, in spite of the -- in industry, there's always some short -- I think you said?

So on your -- the first question on the internalization of packaging. So we do a lot of our capital expenses went into that as we went through the up cycle. And what has happened is the percentage of what we built in-house has gone up. And we constantly look at that as an opportunity to see if we were to make an investment, how quickly would we be able to recover that. And obviously, we will internalize whatever makes sense as long as it has a payback that is reasonable in terms of the time window that we want to get to. And historically, our percentage internal packaging and testing has run about 70-ish percent on packaging. It's running slightly above that today. I think they're non-testing, closer to 80%, 85%, 90%, in that range. But it's purely an economic decision we want to make, which is that it makes more sense and that it will consistently stay at a high utilization by bringing it in-house. On pricing, what I meant is the pricing discipline in times when there's uncertainty can be something that people gravitate towards of, well, how do I win this design on price? And we teach our people and what we have as our system is how to sell on value. And price is one element, but there's so many other pieces of where it's going. It does take a mindset to have a pricing discipline. We have the benefit of a sales force that is not driven by commissions in their regional revenue that they generate. That helps us enormously, making sure that the teams inside don't compete with each other on price. So that their commissions can be driven from region to region. That happens in many, many companies, many of the acquired companies that we have done as well. And so people are looking at how do I optimize for a given situation and retain the discipline on pricing through an economic cycle through, maybe the call that comes from a team for the same customer out of China that comes from, let's say, Boston and how do you make sure that you keep the price whole based on a given design.

We have time for one more in the corner over there.

Can you guys talk a little bit about your data center products there, traction, growth drivers and market share shifts because I think that seems like a very interesting and obviously explosive growth in market?

Yes. So data center is about 18-ish percent of our revenue. data center and computer data center is probably about a percentage less than that, let's say, 17%, 18% or so. A lot of it started in stores network. That's what a big piece came from Microsemi acquisition, but we were also in highly energy-efficient power supplies and some of the other temperature controls and backline controls and all that from the Microchip microcontrollers, et cetera, in -- in more recent times, as the AI servers have taken off, there's been more opportunities in some of those platforms on some of the switching that is required on those platforms, some of the security and route of trust capabilities that are required in these platforms, some of the timing accuracy that is needed on these platforms. And then many new things that we're working on that are intended for the data centers of where things go. So it is -- and we have more detail that we provide that's on our website or otherwise, we're happy to go do it in terms of where the data centers. It is clearly a megatrend that we have for many years have been focused on and the products we build attached and then we also solid-state drive controllers that go into how the memory is being used for high performance as well. So those are all the areas in data center that are key to us in terms of driving growth.

Great. I think that's all we have time for. Thanks, Ganesh. Thanks, everyone.

Thanks, everybody.

Thank you.