Seagate Technology Holdings plc (STX) Earnings Call Transcript & Summary

Okay. Great. It's time. So we'd like to get started. Good morning, everyone. Welcome back to day 2 of the conference. My name is Toshiya Hari, I cover the Semiconductor and Semi Cap Equipment industry, as well as the Hard Disk Drive industry. Very pleased and very honored to kick day 2 off of the Semiconductor portion with Seagate this morning. We have Dr. Dave Mosley, CEO from the company with us. Dave, thank you for coming and supporting the conference.

My pleasure.

I certainly want to spend most of our time going through longer-term, strategic, technology-related questions, but I do want to kick off with the near-term question. A couple of weeks ago, you guys updated your outlook for the September quarter. I think Gianluca did a great job providing context, but we did want to hear sort of the context and the background from you. What you're seeing in the quarter, areas of incremental weakness, thoughts on December, if possible?

Good. Before I begin, I'll be making forward-looking statements. All the risk factors are on our website. Appreciate having me. It's a very interesting time, and we are trying to understand the demand signals that are coming from the entire globe. We've talked a lot about China being well off its prior pace, if you will, for data, and that continues to erode somewhat. As we've gone through the summer, what we've seen is continued lockdowns and just more trepidation of business leaders saying, okay. I think I understand the plans, but I'm not leaning into it at all. And maybe even waiting now past the Party Congress, past Chinese New Year, just wait, wait, wait, so -- and as you wait, there's obviously technology transitions that you may sync up with, so that's been a continued impact. What's a little bit new is the cloud service providers around the world, looking at their inventory positions, they're smart operators. They can't be painted with one brush. They have many different business models. Some of them are very sensitive to enterprise cycles, other consumer cycles. They have great visibility into those things, and I think they've -- and there's still the supply chain overhang, if you will, buying too much of the wrong parts or -- because you have to, or not getting complete kits. So I think everybody is just taking a little bit of a pause here tactically. Long term, we still believe in the growth of data. We've just got to get through this as a company, as an industry. And we've been through these kinds of trying times before, so we're pulling all the levers that we know how to pull unfortunately. It's just to make sure we get through the tactical.

Right. And Dave, just as a follow-up, If you think about sort of the subsegments of your business, nearline via Legacy, perhaps pricing versus volume, is it pretty much across the board that you're seeing incremental weakness? Or is it one over the other? Any kind of additional context that you can provide there?

It is across the board. The only exception to the rule would probably be the highest capacity thing, which we're in the product -- the middle of a product transition to what we call the 20-plus terabyte platform. So we're still leaning into that. But almost all the other or all the other markets are impacted either by consumer spending or by this enterprise inability to satisfy through the supply chain and what's the true backlog out there, and some of that. So I think people are just being cautious around the world. Juxtaposed to the beginning of the pandemic when we had work-from-home, game-from-home, play from home, learn from home, all these other phenomena, those are just not there right now. The consumers aren't helping, and I think the enterprise is starting to look at the consumer, and at least some of the companies, as being really cautious. So very atypical for this time of year, but we're going to have to manage through it.

And Dave, to your point, obviously, a very tricky environment. I'm sure you're getting incremental information day-to-day, week-to-week. And maybe it's a little unfair to ask for your thoughts on December. I know it's early. I don't expect you to give guidance here. But is September, in your view, kind of the trough? And December, you think things could turn higher, or could this be a multi-quarter correction in your view, just given where inventories are and so on and so forth?

Yes, I think it is a multi-quarter correction, but I do think that we'll start coming out of it to some extent. You will see some very, very muted seasonality in some markets. And I think some of these things will be -- that I've just referred to will become clearer over the next few weeks. But it -- this is the time where we're having to really pull back on factory utilization, not only for ourselves and our -- but for our suppliers as well. So -- and just to balance supply and demand properly throughout the remainder of the fiscal year. And I would say we're going to be very conservative on that front to make sure we do it right.

Right. And Dave, on that point, you did talk about making adjustments or responding from a supply side perspective on your earnings call. My guess is you've turned that notch -- up a notch, just given recent developments. But can you share with the audience, again, what your response is on the supply side, both in terms of CapEx and perhaps utilization rate?

That's right. I think we'll continue to dial CapEx down because we just don't need the exabytes right now. And we -- some stuff we can put online just in time, if you will. Probably the most impactful thing is to the people that run our factories or suppliers that are running the factories. I mean, those are factory workers that are suffering from inflation around the world, they have all the other impacts, and we have to make sure that we balance supply and demand. That means turning off the factories. That's not a good situation. So we just have to work through it with everyone and make sure we're ready when the upside finally comes. But right now, that's the highest priority. I think there will be some OpEx trimming. We'll make sure that we look out over the next 1 to 2 years and say these are the investments that are really going to pay off, and be very disciplined on that front as well. I think Gianluca made reference to that in the last conference.

Yes. And then in terms of pricing versus volume in the current environment, you guys, along with your nearest competitor, I think, have done really, really well in terms of controlling supply, managing supply, maintaining discipline throughout the marketplace. Given the current weakness, should we expect pricing to soften a little bit? Or is it primarily on the volume side where you're seeing the correction?

It's primarily volume. I mean, the lesson I've learned time and time again is don't build the wrong stuff, so don't build anything speculatively when you get into periods like this, and then make sure you keep it moving along to turn it back into cash. And we have to not only look at our own inventory, but look at the customers' inventory as well and balance all those things. And so we'll be doing that, but we'll really pull back on the volume first. We are suffering from inflationary impacts just like everyone else, and we have to make sure that we balance ourselves in the middle of that, our suppliers in the middle of that, relative to expectations. So that will need some economic return over time, so we're going to have to be very mindful of that.

Got it. You talked about China as a geo being relatively weak and lacking visibility, if you will, over the next perhaps a couple of months. There have been some geopolitical dynamics in play recently. Not so much to you guys directly, but in terms of high-end data center GPUs and we've been getting a lot of questions from investors about the potential knock-on effects to broader tech, including yourselves. Anything that you've seen from your customers that would indicate ramifications from that very dynamic?

Hard to separate, I think, specifically for us. I mean, we -- to the extent that people are making data center investments that they choose to, for some reason, reduce the data center investment, that may be impacting us indeed. But I think -- overall, I think it's more of just the size of the budget. If bigger pieces of the pie have to go somewhere, then maybe there are smaller pieces of the pie that we're in that, that we're impacted relatively. I think long-term, data is still continuing to grow, so it's going to have its significant wedge there. And I think when people get back into making those investments, they will. So I think some of those things are temporary, and people will find ways to adjust around them.

Got it. One dynamic where we get a bunch of questions is on LTSAs. It's something that provides you better visibility in terms of customer road maps, their long-term demand forecast. I guess the question is how enforceable are LTSAs? Do customers typically stick to LTSAs? What are you seeing today just given the weaker macro environment?

Yes. Typically, they've served us very well, continue to serve us well. So we get -- we could debate the definition of long term, how long term do they really give visibility. But I joked a couple of quarters ago that this is more of a co-planning exercise. Since we're starting material now that may not see the final production for 9 months or 12 months or something like that, then we need to have a pretty good key on-demand signal and the return economics that we're going to get. And I think most customers get that pretty well. What is happening is people are saying, hey, how do I adjust my budget? So inside the LTA, even if they're holding towards it, you start talking about the next LTA or refresh or something. And then you say, we don't want to manage the factories this way. So that's a little bit harder discussion for us and the customers, but we'll get there. I mean, if you look at nearline growth, for example, as a proxy for this, it was huge in '18 and then down during the front of the pandemic. And then huge in '22 again and then down, right? So you just can't run factories that way. And I think the LTAs serve us and our customers as a way of being very predictable to be there when they need us. Typically, in the past, customers have actually pulled above the LTA somewhat, and so it doesn't cover your entire demand. Hopefully, that smoothing function is still going to work for us going forward.

Right. Got it. Shifting gears a little bit into sort of longer-term through-cycle dynamics. I think historically, you guys have talked about nearline demand in terms of exabytes as a 30% to 40% type of growth market. I guess, A, does that still hold in your view? B, in terms of, again, going back to the near term, is it primarily inventory being drawn down on the part of your customers in the U.S., primarily?

Right. So last year was much higher. This year, it will be much lower. So to your point, it's -- 30% to 35% is still a good number? I believe so, but there's a lot of volatility inside of that. Different customers are going through very different things in different geos as well. There's a lot of dynamics right now that are tapping things down. I do think that there's still a TCO proposition around the highest capacity drives, whether they're the 20-plus platforms or even 10 terabytes or something like that. If that happens to be the architecture that you have, it's still move up in capacity, is a good way to use your dollars. So if you're installing something for the next 5 to 10 years, you're installing against that TCO proposition. So that helps us quite a bit. Number of units is a little bit harder, because that will get against the entire budget that people have. Some of that budget may be sucked up on silicon, which is expensive, or some architectural transformation that's going on, network or whatever that the particular individual may be doing. Watch that space because then even if storage is a little depressed now, it will come back later because the data just keeps growing.

Right, right. And how much of this, I guess, softness you're seeing in the near term is supply side related? Meaning tightness, kitting issues on the part of your customers? We hear about networking being a bottleneck, other random sort of penny type silicon components, and maybe that's having a knock-on effect to people like yourself. What are your thoughts on the overall supply chain?

There is still some supply imbalance, but it's abating quickly, especially as demand comes down on the enterprise side. And some of the consumer demand coming down, people pivoted to the enterprise intentionally to try to go after that market. So I think there's still some, but I think in the next couple of quarters, it will probably be gone. That's my estimate. I do think that there's places where people are making investments because they want to, whether it's network investment or a compute investment or something, and memory goes along with that. And if you look at all those dynamics, how much you're investing over there, you say, do I want to make a data investment right now? So -- and generally, and I think people have talked about CapEx this way, it's shrinking a little bit temporarily as well. So all those dynamics will.

Okay. Okay. And then for you guys, specifically from a supply perspective, I think controllers at one point, there was some tightness. Anything that's kind of holding you guys back specifically from a supply perspective?

We talked about more on our systems business and on our drive business, in particular. There were some things on SSDs that held us back. I think most of that stuff is broken free now. Look, there are still some individual small components that are still in very high demand. And I think as the market kind of maybe overreacts to some of the demand signals, the [ demand ] going down, the stuff is sloshing around everywhere, I think it's going to take some time to settle all this out. And it's been a tough year for people to be very predictable on that front. But I do think with some of the demand coming down, inventory positions will be high and people will be able to move parts around and adapt architectures to whatever is available, and I think that's going to figure itself out.

Right. The other topic I want to hit on was your tech road map. It's an area where we get a bunch of questions and we -- being on the sell side, we suffer from he said, she said, and it's hard to sometimes fully appreciate what's going on. Can you remind us where you are from an aerial density perspective? You made some bold comments on HAMR on your most recent earnings call. I'm curious if you've seen any major breakthroughs from a tech perspective, what led you to make some of those more specific comments?

Yes, good. So we're shipping 20-terabyte drives in high volume now. We've talked about that for a couple of quarters. And we can extend the technology via normal, conventional extensions or tricks, SMR and things like that, to other capacity points, so we'll do that with our customers. Some people have advantages for using drives one way and other people in their application space want to use to drive a different way and drives are getting repurposed all the time, so it's a very, very complex deal. I don't think it's great to plant flags rather than just say we have 20-terabyte family. It's fairly well covered for what our customers are going to demand. About a year, 1.5 years ago, some kind of final breakthroughs happened in the labs on the 30-terabyte technologies. And it is around HAMR, heat-assisted magnetic recording. Remember, that's changing the disc properties, the magnetics of the disk, if you will, to be able to pack a bit tighter, and then it's changing the recording head to be able to write stronger and stronger bids. But there's all kinds of other technologies that have to come into bear to make a 30-terabyte drive. There are several mechanical technologies. There's the reader itself to be able to read the data back. The recording channel, which is a deep electronics, signal processing that we have to do. So we've been working on 30 terabytes in development for quite some time, and that's why we made this comment that 30-terabyte drives are going to come out in the late this fiscal year to the middle of next year, and the teams are working very hard in the development cycle. And now we've done development for 40-plus years, so we know how to do that. We need to get the yields up. We need to get the scrap down. We do get the cycle times right, and then we'll go through the product transitions. But it's good news for the industry that there's the capability to get into the 30s. And even the 40s as well, the spin stand demos are really good with this new technology, again, changing the material set on the disc gives us all kinds of opportunities. So that's what we're really focused on. It's 20-terabyte families here, and then 30-terabyte families as soon as we can get there. And then ultimately, we'll work on 40-terabyte families as well.

And maybe it's a little bit early, but can you speak to the customer engagements you're having with 30 terabyte, if any, at this point?

Well, customers have had some of the, I'll say, pre units, not test units, but pre units for some time. And you can imagine that if you're talking about going from 16 to 18 to 20 and doing qualification, pouring a bunch of resource into it, the gains are relatively smaller. But if you're talking about going from 20 to 30, now the gains are relatively bigger. And those are, again, installs that you're going to make for 5 or 7 years in a data center. The TCO proposition for that is great. The amount of bandwidth that you need for that amount of capacity, there's a cost per terabyte, if you will, the power of the footprint, everything else benefits tremendously. So there's a great TCO proposition. And customers are very interested in and then helping us. As a matter of fact, there will be -- just like there always are, there will be new feature sets that come along with it. I mean, the 30-terabyte drives are basically plug and play into the 20-terabyte slots, but there are new feature sets that will make better reliability because we keep working with customers on that and better performance, and things like that. So customers are very excited about it.

Got it. Got it. And in terms of applications, is near line sort of the initial mover in terms of adoption [indiscernible]

Yes. I think so for predictability of that return, we will certainly apply it to the highest capacity points. I think over time, the data that we've seen coming back from the lab allows us to use to apply it to lower capacity points as well with fewer components internal, which helps us address markets that may not be using the highest capacity drive, maybe more edge markets or more price-sensitive markets, and we can get better economics into those markets as well. So that's -- we'll drive for that over the next few years.

And I guess your 30-terabyte product and HAMR more broadly. Could this be a key mover of market share balance going forward? I think historically, you guys and Western Digital, you guys have certainly been larger, and then you've got a very weak #3 player, but it's never been sort of a winner-take-all type of market. But if you launch 30 on time and it's -- you're yielding well and the performance metrics are there, -- could this be a needle mover in terms of market share?

I don't really think about market share that way. It's more -- do I have the cycle times and the yields and the scrap rates and things like that, that I need for the economics? So that's where we're all focused on driving. And then I do think that there are some customers who will say, I don't want that highest capacity drive for that install. But I also think there are other customers who say, I want 2 suppliers, and I'll be able to figure out the world once I understand both people's road maps of those. So I don't really focus on the competition very much. I focus more on what are our customers saying they need from us, and how do we be as predictable as we can with our return on investment.

Got it. Got it. The other question that we get is on sort of hard disk drives versus NAND as a technology. I think most of the substitution on the legacy side has taken place and things have -- if anything, are stabilizing. I guess the long-term question we get is, is NAND a potential risk to near line on a 5 to 10-year basis. Any sort of puts and takes around that from your perspective? Is that the right way thinking of it?

Yes. My general answer to that is no. I do think -- look, I've been in the industry almost 30 years and everybody, there's always something new, some new acronym that comes up and says, we're going to take over hard drives. And I don't think there's a lot of people -- there's not a lot of hard drive people who answer that call all the time, but there's a lot of noise that's generated by people with that accusation. I think if you want to play in a data center at the volume that hard drives do, you have to have performance. You have to -- especially on the right workloads, very intense right workloads. You have to have cost. You have to have predictability of supply, all these things. Tough, tough things to get to for that economics, and I just don't see anything on the horizon that's going to be anywhere close to that. And relative to NAND, NAND is a great technology. It has tons of applications out in the world today. I think -- we actually have an SSD business. We -- it does quite well when we focus on certain applications. I think that's -- that would be my advice to people. To say I'm going to go build NAND fabs to if they cost tens of billions of dollars to get paid back by penny at a time, that's the same kind of economic problem that's always come up with that argument, and I don't see it changing in the next 5 to 10 years.

Right. Got it. Okay. We still have plenty of time. I've got plenty of questions, but I just wanted to pause here for a moment and see if we have any questions from the audience? If you can wait for the mic. Excuse me, can we get a mic please?

Okay. Just given the, I guess, promise of HAMR and also given the backdrop of he current environment right now, does that make you more reluctant or aggressive to spend the CapEx on investing in HAMR?

Yes. So the question is about CapEx relative to HAMR. I think we've been prepping for HAMR for quite some time, so we're mindful of the fact that every new tool we buy, whether it's for capacity or a replacement tool, if you will, because another tool is aging off, is it going to be capable of driving us to where we need to get to in the road map? So I don't think we're going to be starved for any particular tool. In a great environment, would we maybe invest a little faster? I think the answer to that is yes. The transition in the fabs will be more about our ability in our designs to get good cycle times, get good utilization of the tools, OE, if you will, good yields and scrap and things like that. So you're running a fab you want to keep it full, right now, our fabs aren't full. And then you'll fill it up with whatever you can yield economically. So if we had to fill it up with a 20-plus terabyte family or some other family, we'll do that. We like full fabs. And then as HAMR comes in, we'll look at those economics and say, replace back into the fabs, and fabs being so big themselves. But we don't really need incremental CapEx, if you will, for HAMR, especially if we had open space in the fabs.

Any other questions? I'll keep going. Dave, in terms of your legacy business, it's come to a point where, again, the rate of decline is moderating. In the near term, obviously, you've got headwinds. But how do you think about that business, managing that business so that it's no longer a significant headwind to overall financials for the company?

Right. There's a couple of things that play there. So I think the client side, if you will, is really winnowed to almost nothing, notebook and desktop. And I'm not saying we're not making any drives, but they're very, very small volume right now for niche applications. So that's been fairly predictable over the last few years. This front end of COVID work from home drove a little kind of temporary upside. So what really is in legacy now is mission-critical and consumer, and consumer had an upside as well back in those days. Consumer right now, unfortunately, around the world, especially Europe and Americas, is way, way down. So -- and I think some of those are temporary, given what's going on in the macro. Mission-critical is kind of interesting because a lot of people say, well, I can buy this drive versus a mission-critical drive. The reality is there's hundreds of millions of slots out there with SaaS interfaces. There's still a replacement cycle that goes on from time to time with mission-critical drive. So we think mission-critical and consumer can have very long tails. They may go up and down with macro issues, but we think there will be pretty sustainable over time. And so that's what we -- that's the way we watch and think about legacy. It's really not a big part of our investment today. We -- OpEx is virtually nothing. A little bit of customer support, and CapEx is actually being traded off against mass capacity when we see some of these cycles.

Right. And given your strong competitive position, it's pretty much a cash flow generating [indiscernible].

That's right. Yes. I joke sometimes, everybody should have a legacy business that just continues to generate cash without very much investment required. We'll just have to see at what rate is it declining, but I think some of those markets will be -- there will still be volume out there for 4 or 5 years.

We haven't talked about the VIA market. Obviously, pretty choppy. VIA tends to be a volatile business quarter-to-quarter, but what are you seeing there? Obviously, given the dynamics around China? I'm guessing near term, it's tough. But how do you think about that business in longer term?

Yes. VIA is down everywhere. So the VIA is Video and Image Applications, for everyone. It's down everywhere, but I do think that we're on the cusp of some investments over time in edge data. So if you think about smart cities and smart hospitals and big data and AI and understanding consumer behavior, that data has to be parked somewhere. It can't just be processed and then -- I joke sometimes that AI is -- I'll make a decision for you and throw your data away. And just trust me, I made the right decision. I don't think that's going to work everywhere. I mean, in some cases, you have to keep the data for a while, a day, a week, a month. And that's where some of those VIA applications will just grow like crazy. I think right now, at the front end of the pandemic, people couldn't get on-site, so there wasn't a lot of on-site edge installation going on. Now people can, but I think they're so cautious about budgets around the world that the market's just soft. And China, in particular, which has made a lot of those investments, is soft as well. But Europe and the Americas is soft. In around the world, most of the VIA markets are what I call the white van business. It's the person who goes and kits the gear, and then goes and does the install. I think what we're starting to see a little bit, which is an optimistic sign, is people are saying, "hey, I can install a system for this building, hospital, hotel, whatever it is. But I need to link that up with other things and start to use the data a little bit broader." And so those are the kinds of trends that optimistically, we think will start to break free the markets.

Okay. So the drivers are there, but the economy is the one that's pulling that back.

That's right.

Understood. In terms of gross margins, in the near term, clearly, you've got inflation, cost inflation and macro headwinds. Medium to long term, I guess the view that many people have, including myself, is it's a very concentrated market from a competitive standpoint. You guys build very high-end complex products that takes months, quarters to build. Why couldn't gross margins be higher? And what are your thoughts on supply side discipline that you see across the marketplace? The value that you can capture, again, longer term, perhaps?

I completely agree with what you just said. You said it really well. Keeping supply imbalance proper in a very volatile world is the hard nut. I mean, if you think about our industry 10 years ago, we were at the peak of client server, that was -- we were making notebook drives for everybody on the planet that was trying to get on the Internet for the first time. And a lot of the enterprise gear that was still on-prem. The pivot to cloud and data going to the cloud has been incredible, but it's also dramatically impacted our factory footprints from low cost, high volume, good enough quality for your notebook to good enough quality to run 5 to 7 years in a data center. No pit stops, right? And our suppliers have been through a lot there as well. So the way I think about gross margins is that's our target for return, and we have to balance supply and demand exactly against that target. When we get in quarter, then you're chasing business that hopefully is not dilutive to yourself, but you need to turn cash and things like that. But the way we're going to install CapEx and the way we're going to build products and things like that is against those gross margin targets. And I do think that as data continues to grow over the next 10 to 20 years, there's opportunity for us to get the footprint reset after the client server peak now and into this new peak that we're going into with the cloud and the edge, that will hopefully get better supply-demand balance predictability.

Right. And is it fair to say that you're seeing discipline from your competitors as well? If it's an industry-wide kind of dynamics that's improving?

Yes. I don't really pay very much attention. We can't really see it all the time. But I would say that everybody is probably going through the same kinds of demand crunch right now, and probably having to do is be as disciplined as they can on the supply side as well. It's just some of these things are happening very quickly. Some of the inflationary aspects, whether it's freight rates last year or wages or precious metals or -- whatever is going in electronics components, whatever is going up, it's really hard to balance all those things together. So I think that's why you have to set your footprint and kind of retarget all the time. So I think the industry needs to get paid better. The suppliers need to get paid better for what they're delivering to the world, and I think that's why we have to enforce these disciplines.

Got it. And then from a spending perspective, at the top, you talked about being a little bit more disciplined, and perhaps scrutinizing some items more so than 3, 6 months ago. I guess, what's sacred? I'm guessing R&D, but what areas are available for you to potentially save on OpEx going forward, just given the environment?

Yes. We have aerial density investments that will protect some -- to some extent. There's always things that are speculative that are going on there because we're trying things for 50 terabytes and 60 terabytes, the things that may not be panning out too, so we can always pull back a little bit. Drive development has become very honed. We only have a couple of different drive types now whereas once upon a time, we had notebook drives, even iPod drives and mission-critical drives and nearline drives and desktop drives, so a lot of that consolidation has helped us. And we actually do have a couple of different drive types. Someone will ask the question, do we want to refresh that as much as we can? So we can pull back on the R&D side, even a little bit. There are some other projects that we have that look out more like 10 years to say, this is the kind of market development we think or this is a technology interface development, for example, electronics development, and maybe we want to put that on hold for now. So I think we can -- I think we did last year -- or sorry, in the front of the pandemic, we did pull the reins back and I think we're going to do that again. Just wait and see kind of what next year brings on that front.

Got it. And then from a capital allocation perspective, I think you guys have been pretty aggressive in buying back stock. And I think investors have appreciated that it's shown up in your relative stock price performance. As you look forward for the next 12 months with the macro doing what it's doing, how are you thinking about investments in the business, returning capital to shareholders? What sort of the debate [ and thoughts ] you did?

Yes. No real policy change. Although to your point, the kind of forward look has changed quite a bit, but no policy change where it's still going to have the dividend, protect the dividend. We'll look at buybacks opportunistically as we always do, and we'll also just defend our cash through this period of time, be very, very conscious of that. That's what we did a few years ago. As an example, I think our capital allocation policy has served us pretty well over the last few years, and I don't see any reason to change it right now.

Got it. Okay. I want to pause here again and see if anyone has any questions. Good. I guess, Dave, in the last couple of minutes, I want to give you the opportunity to speak to anything that we may have missed. I know you spend a good amount of time with investors, but as a group, anything that we overlook or underestimate, under appreciate about your markets, about your tech, your road map?

No, I think good questions today. So I think you've got most of it from -- from my perspective, it's an interesting time to run big factories and big supply chains in the world. There's a lot of moving parts. Literally, we joke about that at Seagate all the time, having to make tens of millions of disk drives.

And how many parts in it?

Yes, it's hundreds of parts in each, and a lot of the specialty mechanical parts. And so we have a lot of factories to take care of. A huge number of people, tools, to make sure that everything is aligned. And I'm cognizant of our responsibility in that to not only make sure it gets through tough times and do our part, but also to make sure that the customers know that. What we'd really like to do is build a sustainable business going forward, and these up and down cycles are a little problematic. When we saw that huge peak of client server, we had to come all the way down and transition to mobile -- to mobile cloud. There's another transition coming at some point, so getting that right is what I spend a lot of my time focused on. Getting your people ready for those changes that they'll have to go through professionally, getting the suppliers healthy so they're ready to respond at the right levels as well. And last year was good. I think everybody benefited this year. There's a little bit more trepidation around the world, and we've just got to make sure we get both hands on the wheel, so.

Awesome. Great. Thank you. Thanks so much for the insight.

Yes. Thanks very much. Thank you everyone.