Micron Technology, Inc. (MU) Earnings Call Transcript & Summary

Hello. Thank you for standing by, and welcome to Micron's post earnings analyst call. [Operator Instructions]. Please be advised that today's conference may be recorded. [Operator Instructions] I would now hand the conference over to your speaker today, Farhan Ahmad, Vice President of Investor Relations. Please go ahead.

Thank you, and welcome to Micron Technology's sell-side analyst callback. On the call with me today are Sumit Sadana, Micron's Chief Business Officer; Manish Bhatia, Micron's EVP of Global Operation; and Dave Zinsner, Chief Financial Officer. Today's call will be approximately 45 minutes in length. As a reminder, the matters we will be discussing today include forward-looking statements. These forward-looking statements are subject to risks and uncertainties that may cause actual results to differ materially from the statements made today. We refer you to the documents we filed with the SEC, specifically our most recent Form 10-K and 10-Q for a discussion of risks that may affect our future results. Although we believe that the expectations reflected in the forward-looking statements are reasonable, we cannot guarantee future results, level of activity, performance or achievements. We are under no duty to update any of the forward-looking statements after today's date to conform these statements to actual results. Operator, you can now open the call for question and answer.

[Operator Instructions] Our first question comes from Ambrish Srivastava with BMO.

Dave, I actually had a clarification on the supply/demand side. When you were talking about -- when you get the days of inventory guide for the fiscal year, you mentioned that Micron would be tight. And I think last earnings call, you had said for fiscal '22, you would be below the supply that the industry would be turning out. So I just wanted to make sure I understood that. Would you be seeding some bit share if you would be tight and the industry would be healthy? I just want to make sure I had clarity on that.

Yes. So we expect to be in line with the, in terms of shipments with the industry. And you're right, the industry should be in balance. But all I'm saying is we will have built up some inventory. We're already up to 103 days this quarter. We're likely to be higher than that in the second fiscal quarter. And then we'll be leading into that inventory within the third and fourth quarter. So we're staging our build plans to be able to manage through the demand for the calendar year.

Got it. And then a follow-up is on free cash flow. You've been pretty positive about the ability to generate. I think you've characterized it as healthy free cash flow. And you've given a backdrop of the industry being healthy demand driver. I just wanted to drill a little bit more. Is the underpinning of the confidence is also on a structurally more profitable company, i.e., more profitable product portfolio combined with the LTAs and leading to better visibility that you have? Is that kind of factoring in besides the industry dynamic that is leading you to be so confident about the free cash flow for the year?

Good question. I think there are several dynamics. First of all, we come into the first fiscal quarter generating good free cash flow. We generated $670 million or so of free cash flow. So that's part of the confidence. Also, when you look at our margins for the first fiscal quarter, which was perhaps a tougher quarter from a revenue perspective given the -- when you see shortages -- our shortages in the PC industry, we still generated 35% operating margin. So we are obviously a different company than we were 5 years ago in terms of our ability to generate good profitability through cycle. And so that tested a lot to do with the confidence. When you look into the second fiscal quarter, and take the midpoint of the guidance, our operating margins are about 33%, so also very healthy. We do expect to generate free cash flow again in the second fiscal quarter. And then as Sanjay mentioned, we go into the third and fourth fiscal quarter with strong demand drivers, which gives us confidence. Now granted, we're going to invest $11 billion to $12 billion of that -- of cash flow from operations back into the business. But even with that, we think we'll have a very good year from a free cash flow perspective.

Our next question comes from Aaron Rakers with Wells Fargo.

Congrats on the results. I want to go back to the inventory discussion, Dave. I know you just clarified that you expect to exit the year or maybe fiscal year with less than 100 days. How would you characterize, in that context, the company's ability or willingness to kind of leverage that inventory tightness on your own side from a pricing perspective? Have you seen situations where you stepped away from some price discussions? And how do you think that, that materializes that inventory days kind of trends lower into the back half of the fiscal year? And then I have a follow-up.

That's a good question. We're always willing to hold inventory. We don't think we're capturing the right value for the products. We allowed our inventory to go up into the 130s. I think at one point, we were even as low as -- or even as high as the low 140s in terms of days of inventory. And we feel very comfortable going to those levels, particularly as cost declines have become more modest in particular Instagram. And that allows us to have more flexibility in terms of holding inventory. And we as you point, we test the pricing all the time. Sumit can probably talk a little bit more about this in a second, but we are testing the pricing all the time with customers. I think this circumstance where we are in terms of inventory is that we won't be in that position. We won't see inventory trending up to those levels. And when it's down below 100, it's actually a fairly lean level for us based on how things have played out over the last 12 months. I think ideally, we'd like to be closer to the high end of our optimal range of 105 days really to effectively operate. So it's going to be tight. But yes, if we need to, we can always hold inventory if we don't like what the returns are for those products. But I don't know Sumit, if you have anything to add?

Yes. I think, Dave, you mentioned this well. I mean -- and I agree with the sentiment that we feel like being closer to the higher end of the inventory range. It's just better place to be in this type of an environment where lead times are long for all sorts of parts that we need to procure, our customers need to procure and so it becomes more challenging to respond to mix changes if the inventory is very low. And so I think we have definitely been having discussions with customers when pricing doesn't meet our targets and what we think is the value we are providing in those products to our customers. We would much rather hold inventory. But even then, we have found that there are other customers who are willing to pay up and take the product that we want to step back from selling to a customer who may have more aggressive views on pricing. And that's part of the reason why our gross margin performance has been pretty robust, and we continue to focus on ensuring we get the right value for our products.

That's great. And then on a follow-up question, I'm just curious, there's going to be an increasing focus on DDR5 as we move through calendar '22, along with some other architectural things going on in the data center side. Just when I think about a server with DDR5 looking into the back half of the calendar year relative to today with DDR4, how would you characterize the amount of gigabytes of DRAM capacity expansion we would expect to see as that transition starts to play out on a per server basis?

Yes. I can talk to that. So really, the server transition to DDR5 is just starting in late 2022. And it's going to be a multiyear trend. And what we will expect to see is roughly a 20% increase in overall average capacity per server over 20% increase. And on top of that, the whole industry is going to move to higher density of memory die in order to support larger module capacities of DRAM as well. And the reason this is important is because the core count in these newer architectures is going up pretty significantly, sometimes 30%, 40% increase in core count. And so the memory per core that is needed is going to require those higher-capacity DDR modules. And consequently, DDR5 additionally helps by increasing the bandwidth, increasing the performance of the memory module. And the average capacity is the other important aspect that our customers will be looking for beyond the improved performances itself. So the average capacity trend continues to improve the overall -- if you look at some of the AI and machine learning specific servers that are purpose built for AI and machine learning type of applications, you will see that DRAM and NAND content is going to be 50%, 60% of the BOM cost. The single largest portion of the BOM is going to be just DRAM and NAND. And everything else, including CPU, GPU, networking, everything will be 40%, 45%. So it's a very big step-up in the cloud for DDR5.

Our next question comes from Harlan Sur with JPMorgan.

Your data center business was up fairly strongly. It's up 70% year-over-year. Was it up sequentially? And you guys talk about enterprise and data center being the largest market for memory and storage. So is data center and enterprise, the largest part of your DRAM business now even bigger than your mobile DRAM business?

Sumit, do you want to take that?

Sure. Yes. So as we look at it from a quarter-to-quarter trend perspective, definitely, we do have some mix changes in terms of the different segments because when we have strong demand, it does become a discussion of how much of the demand goes into which segment. But generally, it is definitely true that the data center market is the biggest market for our TAM. So when we look at memory and store together, the data center market is the biggest portion of the TAM. And it is also one of the fastest growing. Now the fastest-growing segment is automotive and industrial, but a close second beyond that is data centers, particularly if you include networking and the networking is growing very fast as well. So data center is both the biggest portion and one of the fastest-growing segments. And within the data center, we look at cloud and hyperscaler type of customers growing faster than the traditional OEM customers, right? So with that having been said, as we look forward to 2022 and beyond, I just want to call out that part of the confidence that we have in terms of our second part of the fiscal year and second part of the calendar year growth is really driven by a very robust capital spending plans from some of these cloud companies. So if you look at how they have been talking about their investments and you see some of their plans, they have a pretty substantial step-up of investment in cloud infrastructure and server infrastructure in fiscal '22. And you, for example, just look at Meta as an example, they said they spent $19 billion -- they're going to be spending $19 billion CapEx in fiscal -- sorry, in calendar '21, but in '22, they are expecting that number to be in the $29 billion to $32 billion range. So very material step-up. And that's part of the reason why we feel like that segment is going to grow strongly for us. And we have a very strong value proposition in that segment. Our product portfolio there is really strong. We have leadership in DDR5. We have leadership in quality, 2 out of 3 customers rank us #1 in quality, and so they prefer that in a big way. And then our data center SSDs have been -- they're looking forward to a really good year there because we just introduced our vertically integrated platform there after years of working on it. So another very exciting part of our portfolio transformation. And then the last piece I'll mention is you have seen a lot of graphics growth in the data center and our graphics memory with the G6X -- GDDR6X, we've the world's fastest performing graphics memory bar none. And we have a great partnership with NVIDIA to drive that for further gains as well. So lot of legs to the story in many different aspects and a really exciting time for us for our product portfolio in the data center.

Yes. And I appreciate the insights and maybe sort of as a follow-up because you brought this up. So it looks like I've been tracking your guys' SSD business. It looks like even before the launch of the 7400 platform, I think your enterprise and data center as this deep business is about the same size as your client SSD business. Is that about right? And then, how do you see momentum going into next year? You guys talked about introducing the new 7400 data center platform in October, strong ramp in fiscal Q2. Is this with multiple Cloud Titans?

Sorry, I didn't catch the last word of your question. Is this multiple what, sorry?

Yes. Is the 7,400 ramp in Q2, is that being driven by multiple Cloud Titans ramping?

Yes. So I think there is no doubt that if you look at our high-value NAND mix we have been on a very steep trajectory to improve the percent of our NAND output that is being sold as high-value solutions. And in that increasing our mix of SSDs is very critical, we have a really strong portfolio for both client SSD that go into PCs as well as data centers. So we have a strong product cycle going on both the 146 of the -- sorry, 176-layer NAND-based SSDs for client as well as our vertically integrated platform for data center which we have just launched. So we are seeing interest from all sized customers, the cloud companies as well as traditional OEMs as well as channel partners for this vertically integrated platform because they believe it's all of the early testing they have done, they find that product to be very attractive. So between this product and future generations of this product, we are pretty confident that we'll be able to build a pretty strong momentum. And the data center SSD business, as you know, is the most profitable portion of the NAND portfolio, and so as we finally start to ramp this over the course of next several quarters, it should be a tailwind for us in our financial performance. So I feel pretty good that over the next year and 2, we will end up with much improved share in data center SSDs than what we have had in the past couple of years.

Our next question comes from Srini Pajjuri with SMBC.

Dave, on the balance sheet inventory question. Just based on your guidance, it seems like you're guiding for bits to be kind of flattish in the second quarter, fiscal second quarter. So my question is you said your inventory still go up. So I'm just wondering, even if bits are flat, do you think your inventory actually goes up on that?

Well, we're not expressly guiding bits for the second fiscal quarter. But I would say I'd stand by my comment that we expect our days of inventory in our absolute inventory to be up a bit in the second fiscal quarter.

Okay. Fair enough. And then there was a comment about LPDDR5 penetration in the PC market being about 20%. So obviously, you're really well positioned in that market. And -- but my question is, are there any implications that we should be aware of for either pricing or margins or anything else as LPDDR5 becomes a bigger portion of the PC market?

I'll let Sumit comment on that.

Yes. No, I think that's more of a comment we provided to just provide some color on some fundamental shifts that were taking place in the client market and the PC market. We are very strong in both DDR compute products as well as low power. And some of our low power products have really good capabilities that our customers appreciate, but we have world-class quality in both and very good performance in both. And so not really much of a financial comment there as much as just color on the industry trends.

Our next question comes from Tom O'Malley with Barclays.

Dave, let me try to attack the big question in a different way. So previously, you had talked about difficulty growing bits in the first half and then any upside would come with the transition to 1-alpha. Obviously, to get to your target for the year, it implies a substantial back half ramp. But can you talk about how that transition happened a bit better than expected? And given that, do you think that it's that assumption to maybe see some of that bit growth come in a little earlier in the year, just given the amount of strength you see for the entire calendar year.

So good question. Let me let Manish take that since he's in charge of ramping it.

Tom, thanks for asking. And yes, no, 1-alpha and 176 are actually doing better. They're ramping really well better than maybe we expected a quarter ago in terms of how they're yielding, how much the productivity we're getting out of those tools and that are dedicated there. And so you're asking your question specifically about DRAM, I think we made a comment that we now have 176 players is now the majority of our NAND production and that 1-alpha and 1Z combined is, as we said, 50% of our total DRAM output. So and I would expect that 1-alpha as we continue to ramp it to actually both 1-alpha and 176 to continue to ramp and grow as a portion of our mix as we go through the year. And as Dave mentioned about, we could see inventory growing a little bit in this next quarter, but that just positions us well for a strong second half of the year ramp and getting more of the 1-alpha and 176 cost capability into our P&L and the results that dropped to the bottom line in the second half of the year.

Yes, I would go so far as to say, we have to build this inventory ahead of our strong demand expectations for the back half of the fiscal year. Otherwise, we will be challenged. So it was actually part of the strategy really.

That's helpful. And then in your commentary, and then I think in the deck as well, you talked about some strategic supply agreements that you guys have entered into, just to help on the supply side. You mentioned UMC, but can you talk about any other parts or strategic agreements that you've got into that kind of help with that supply as the year goes along?

Yes. So Manish, you probably -- strategic supply agreements?

Yes. So for example, we've said publicly that we'd entered into a strategic supply agreement with UMC recently to be able to provide a supply of controllers that are important, in particular, to our automotive business and some of the other businesses in our NAND portfolio. And we have other agreements on other components. The analog space is definitely an area that has been challenging for the entire industry. We're also doing things with back-end materials like substrates and other materials that go into assembly and test processes to make sure that we have sufficient supply to be able to meet our plans for the rest of the year and meet that supply with the varying mix of supply that Sumit was referring to -- of demand that Sumit was referring to earlier.

Our next question comes from Karl Ackerman with Cowen.

Clarification and a question. Dave, first one's for you. Would I be wrong to conclude your November results witnessed DRAM costs improved, low singles on a year-over-year basis and NAND improved low teens on a year-over-year basis. And I guess I'll just ask the follow-up question for Sumit or Manish. I guess I was a bit surprised that Sanjay indicated, you have mid-single-digit share in automotive, and that would certainly lag your overall share in the market. I'm curious why is that? And maybe what -- more importantly, what steps have you taken or can you take to improve your competitive position within this market that is certainly going to be a growth driver, not just for this year, but the next couple of years?

We should probably address the second part first, Sumit, do you want to, I think there's a misunderstanding. Do you want to explain that?

I'll add to the comments that Sanjay said on the call. So I think what Sanjay implied was that our revenue, that exposure, maybe like auto and industrial is 10% with no point, did he say that we have a mid-single-digit share. We -- in the call in the prepared remarks, we said that we have market share leadership and we have said it many times. So -- and with that, I'll let Sumit comment more on our proposition.

Yes. I think there is some level of communication or misunderstanding. Just to be clear, on the automotive side, Micron is #1 in the world in automotive share. In fact, our automotive share is so high that the next 3 competitors combined have less share than we do. So we are #1 by a large, large margin. And if you also look at the automotive plus industrial share, we are #1 in the world in automotive and industrial together. And independently in industrial as well, we have an extraordinarily high share, much higher in both cases for industrial and automotive than our supply share in the industry. So you have seen already that we mentioned that our industrial IoT growth in Q1 was over 80%, year-on-year, our automotive growth in spite of flat unit shipments and the car companies, our automotive growth was very strong. We've been setting records in quarterly performance in automotive in spite of our end customers being severely constrained in their unit shipments. And all of that is testimony to be strong amount of growth of average content. And so we have been investing in automotive for decades, and we have a long history of high-quality product assurance of supply through multiple technology generations and a lot of good capabilities that we bring to that market, long history of support to our customers there. So we are very, very well positioned in automotive. And since automotive and industrial are the fastest-growing segments in memory and storage over the next decade, this #1 share, far higher than our supply share, should position us really well to profit from that.

And then on the cost side, in DRAM, so maybe just level set, we think that cost declines in the industry for DRAM are mid- to high single digits. I'd say in the first quarter, our year-over-year cost declines were more on the lower end of that. On the NAND front, we think cost declines generally run in the mid-teens, and we were a bit below that in terms of our cost decline. It's partly influenced by mix. And also, I think we just have a bigger chunk of expenses associated with COVID mitigation and so forth that's impacting us in the first quarter that we weren't seeing nearly that level in the first quarter of last year. So that's been somewhat of a headwind to the cost declines year-over-year.

Our next question comes from Chris Caso with Raymond James.

I guess first question is what you think about customer inventories right now or any areas you think may be higher or lower that you'd wish to call out? And then I guess with some of the supply tightness that you expect in the second half of the year, I guess, might your customers be expecting that as well? And might they have some incentive to take some inventory just as you guys are building some inventory now.

Okay. So Sumit?

Yes. So in terms of inventory at different customers, the -- there are, I would say, the overall inventory picture is next picture. There are some customers who have below target inventories, and there are some who have been carrying inventory a little bit higher than their target. One example of the latter is phone companies, smartphone companies that are based in China have generally a little bit higher level of inventory than their target. Now the other aspect of this whole inventory issue is that customers have had a tough time with inventory, getting enough parts, having adequate supply of semiconductor inventory. So over the last 18 months, they've experienced a lot of challenges. And so their overall target levels of inventory could be higher strategically. And for extended periods of time, we don't know. So that's one aspect. The other is always the geopolitical aspect of what happens with inventory. So some companies have chosen to have high levels of inventory to manage all of the supply chain risks that occurred due to geopolitical or other aspects. So I think with all of that said, we feel like the demand environment is still very, very good and very robust. And the overall level of end demand is still not being fully reflected on the memory and storage side because there are shortages in the semiconductor supply chain that are constraining demand. And as that demand -- as that supply improves through calendar '22, we should see more of that inventory consumed, and we should also see more of that demand to show up into memory and storage industry broadly. So that's sort of a quick summary, I just wanted to provide.

Okay. That's very helpful. As a follow-up, in your prepared comments, you talked about your expectations for a healthy supply-demand balance in calendar '22, especially in DRAM. Maybe you could expand a little more on those comments as it pertains to industry supply growth and kind of what you're expecting there. And with that, the equipment lead times are very long. So to what extent do you think that those equipment lead times are serving to constrain the capacity additions even if some of your competitors wanted to add more.

Manish, you want to take that?

Sure. Thanks, yes. Thanks, Chris. So I think there's a few different factors contributing. I mean we talked about the demand side a fair bit looking all the end markets looking healthy, except for this air pocket, we just hit in PCs, and now we CPCs improve. So with all of them coming together, we see strong demand for the second half. Now in terms of supply, there's a couple of things to think about. First, keep in mind that last year was a very strong year for inventory depletion at all of all competitors in the industry, right? So the comparison as we look ahead to '22 is based on a very high inventory depletion in addition to natural new supply output. So that's one thing. The second is that technology transitions are continuing to get harder and harder for everyone. And so we're not getting as much big growth out of each station as an industry as we used to. And then the third one is the part that you asked about, which is the lead times for equipment. And I think you've heard both from equipment -- sorry from all competitors that we are actually being prudent with our CapEx spending. For example, we said that we're investing less in DRAM manufacturing in fiscal year '22 than in fiscal year '21. And then you've heard equipment vendors say similar things about lower targeted DRAM investment from their -- from all of their customers together. And then the lead times with regard to equipment, our -- that extension, it just kind of keeps an overall lid on how much the growth -- when new capacity could come online. And the extension there does kind of help make sure that the supply-demand balance remains stable, through '22. And I mean we're even seeing, having to work with our suppliers into '23 now as well in order to make sure that we're securing all the slots that we need. And I'm sure the other competitors are doing the same. That's just the nature of the equipment industry dynamics right now.

Our next question comes from Raji Gill with Needham & Company.

I'm sorry about that, I was on mute. So a question on the gross margin. The margin is guided to be about 46% down sequentially. That's in part due to seasonality. But I wanted to get a little more clarity on kind of the near-term outlook for margins. And as we progress throughout the fiscal year '22 at a high level, is it fair to assume that the margins should start to inflect higher as you see kind of more moderate ASP deceleration as well as more of a better cost structure? So just thoughts on how you're thinking about margins as we kind of ramp into the stronger fiscal second half and near term as well.

Yes. So without...

We tend not to obviously comment on margins beyond the first -- one fiscal quarter out. I guess what I would say is there's some things that are going in our favor. One, we'll be continuing to ramp, 1-alpha DRAM and 176-layer NAND, they have good cost structure, the more they come in the mix, the better, I think we are from a cost perspective. And two, as Sanjay mentioned, we have the introduction of the 7400 SSDs. But just in general, the move both in DRAM and NAND to higher-value products, I think, is beneficial to us in terms of driving better profitability, better profits and so forth. So those 2 things are in our favor. Pricing, we'll have to see how that plays out on the, the only negative thing I think we have in front of us from a cost perspective is really just the costs associated with managing the supply chain. That's coming in the form of COVID, but also just in the form of just lots of pressure on the supply chain that we're experiencing incremental cost, and that is a headwind. I think maybe the only other thing to say is that these margins that we're operating in are quite healthy. And if demand is strong like we believe it will be in the back half of the fiscal year and into the back half of the calendar year, I think we have -- we're in a good place. The only other thing I'd add is, which I failed to mention before, but the expenses we're experiencing, these headwinds that we're experiencing, our competitors are experiencing, the entire semiconductor industry is experiencing, so this won't be something that's unique to Micron in any way. This is something that -- one that everyone is seeing.

And for my follow-up, on the 5G, the smartphone business, the mobile business, I'm wondering why you admit you're characterizing the overall end market as we close out the year. And then as we go into next year, you talked about I think nearly 700 million of 5G smartphones. So obviously, 5G as a percentage of the overall handset market is increasing. So I wanted to get your thoughts on kind of some of the DRAM or the NAND content drivers in 5G as you go into next year. Is there going to be higher dollar content there? Or is it just more about that the unit as a percentage of the overall end market is going to be representative of higher content per phone?

I'll let Sumit address that.

Yes. I think the 700 million plus unit view of calendar '22 gives us roughly 40% plus growth of units from calendar '21 to calendar '22, so that's a positive. And the reason it's positive is because the 5G phones have higher average capacities of both DRAM and NAND. So DRAM typically 50% higher average capacity and NAND is almost double average capacity. If you see some of the phone models from Apple and Samsung, you'll see the strong improvements they have made from 4G to 5G phones or LTE phones to 5G phones. So that 5G trend is very good. And keep in mind that these are just being driven out of existing known applications with a lot of AI usage when you take pictures that's happening in the background. So that's driving a lot of content of DRAM. And when you save in 4K video, et cetera, or even draw data on the photo screen captured now for post processing of these images, a lot of that requires large file sizes. And so these are just known existing applications. As you know, at the start of any new technology when new applications come up that take advantage of 5G overall bandwidth increase, lower levels of latency, et cetera. So you see all of these being rolled out in the years ahead, including augmented reality type of applications for phones, including more usage of phones for gaming. Then you will see the average capacity trend continue beyond the initial burst of this going from 4G to 5G. So that's how we see the progression happening driven by applications, and that is still very early in the 5G cycle. That's going to be a decade-long trend, and we are just right now seeing the benefit of just the jump from 4G to 5G and then future average capacity growth beyond that 2 applications.

Our next question comes from Steven Fox with Fox Advisors.

Just following up on all the mix commentary for the second half of the fiscal year. Dave, if we just isolate around mix, it seems like with data center and maybe even content on mobile and then auto that, the mix story in the second half is pretty strong for gross margins. Can you just explain or just maybe walk that back a little bit, what would be the major offsets to thinking you get decent margin -- gross margin expansion in the second half? And then I had a quick follow-up.

We're not going to predict gross margins, but clearly, that will be beneficial. And as I said, clearly, across the technology, cost structure will be beneficial. And then the offsetting negative would generally be the -- from a cost perspective, would be these supply chain costs that we'll experience. But we're not prepared to guide for a third or fourth fiscal quarter. I think all I can say is stay tuned for the 2Q earnings call, and we'll give you guidance as to where we think things are going to shape up to in the third fiscal quarter.

Fair enough. And then just real quick on the improvements you're seeing in the supply chain. Like how would you describe those in terms of how they start to layer in? Is it pretty straight line starting after this quarter? Is it we're still -- do you get a bulk of it as you get to the fourth quarter? Any other color around that would be helpful.

Sorry, the improvements in supply chain is that what you're saying?

Yes, yes, the improvements in your own supply chain that you talked about.

Sure. So I think as we've gone through this year, we've seen shortages across a few different end components. I mentioned analog. I mentioned that's primarily -- the primary driver of some other logic components. And we've been able to, I think, mitigate most of those pretty well, and we've taken actions across those areas to make sure that we will see improving supply as we head into calendar year '22. Having said that, this kind of environment is such that there are -- while we address certain issues, new areas of challenge pop up, and we're working aggressively to make sure that we can build inventory on those -- across all the different components that we source so that we're able to address all the demand that we see for next year. So we're in a pretty dynamic situation, but with these long-term strategic agreements, we'll gradually build inventory on the parts that we need such so we can meet the demand that we have.

Thank you. This concludes today's conference call. Thank you for participating. You may now disconnect.