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

Good morning, and thank you for attending JPMorgan's 19th Annual Technology Investor Forum here at the Virtual Consumer Electronics Show. My name is Harlan Sur, semiconductor and semiconductor capital equipment analyst. Very pleased to have Sumit Sadana, Micron's Executive Vice President and Chief Business Officer; and Dave Zinsner, Chief Financial Officer at Micron, both here with us today. I've asked Dave to give us an overview of the just reported November quarter and February quarter outlook, then we'll launch into the Q&A. So gentlemen, thank you for joining us this morning and let me hand it over to you, Dave.

Thanks, Harlan. And obviously, thanks for having us. Sumit and I, of course, will likely make forward-looking statements and encourage all investors to look at our risk factors in our SEC filings that -- including the most recent 10-Q that we filed last week. Okay. So I will be really brief because I'm sure most of you have read the transcript or were listening in on the call. So pretty good quarter for us. Originally coming into the quarter, we thought revenue would be somewhere in the $5 billion to $5.4 billion range. Sanjay, towards the end of the quarter updated it, thought it might be between $5.7 billion and $5.75 billion. We actually even slightly exceeded that number, $5.77 billion, at the end of the day, which was obviously a great result given all the uncertainty out there. I'd say demand generally across a lot of different markets was particularly strong both in DRAM and NAND. Also, pricing was better than we expected in both DRAM and NAND. And so not only did that help on the revenue side, but also on the gross margin side we ended up -- originally, we were thinking the likelihood was -- would be below 30% gross margin. So almost like in the ZIP code of 31% gross margin is where we ended up, so pretty pleased with that. On the earnings side, we did $0.78, so well in excess. The midpoint of our guidance originally was about $0.47. And we updated it to think it might be in the low 70s. It turned out to be in the high 70s. $0.02 of that was some unusual events related to Micron Ventures. But for the most part, it was all due to really good revenue strength, good pricing. And we actually did a little bit better than expected on the cost side as well. As we look into the second fiscal quarter, things continue to look good for us. I think most of the markets, I'm sure you'll get into questions and Sumit will give you some details around some of the markets that still haven't quite caught up to the strength. But for the most part, the demand has been pretty broad across a lot of different markets. Right now, we're thinking revenue $5.6 billion to $6 billion. So at the midpoint, roughly aligned with the first quarter. But typically, the second quarter is a seasonally weak quarter for us. So we're overachieving relative to the normal seasonal patterns. Gross margins, happy to see that we think gross margins will be 30% to 32%. So we never -- in this second little downturn, we never actually went into the 20s. So I think that was good execution by the team. And then earnings, we're generally thinking $0.75, plus or minus $0.07. So roughly kind of similar earnings levels to what we saw in the first quarter. And then I think before I kick it back to you, Harlan, the one other thing, I just want to point out that one of the bigger things that we wanted to get right was the inventory levels. They had been elevated. We were working them down early in '20. And then, of course, with COVID, they started to come back up in terms of days. And last quarter, they were up to 135 days of inventory. And we cut off about 10 days of that level just in the first quarter, so down to 125 days. So we're quickly getting those inventory levels down into our optimal state, which revised with the accounting changes we're going to make, would be somewhere between 95 and 105 days. And we think we'll make pretty good improvement in the second quarter as well on those days. So I think that sets us up for a very strong, not only fiscal '21, but calendar '21 as we proceed. So that's just the high levels. And with that, I'll kick it back to you, Harlan, and you can ask the detailed questions.

Yes. Absolutely. Thanks for those -- thanks for the opening remarks. So heading into calendar 2021, I mean, our view is the semi industry is going to grow 8% to 10% this year. Our economists are forecasting 5% global GDP. Cloud spending looking like it's going to grow 20%. 5G smartphone is going to grow more than 2x. 5G infrastructure spending is going to grow about 35%, automotive production, up 15%. So feels like we're stepping into a year that is going to be quite strong. I'm curious from Micron's perspective, what end markets do you expect to exhibit the strongest demand as it relates to driving your products? And what areas do you still think are going to be relatively weak in calendar '21?

Yes. Harlan, thanks for having us. I think from a Micron perspective, it's tough to choose a segment of the market that's going to be weak for 2021. We are seeing very broad-based strength across multiple segments of the market. You just rattled off a few of those. And you mentioned global GDP and the semiconductor overall industry growth. Clearly, we see semiconductors growing faster, much faster than global GDP overall, but we also feel that memory and storage will be the fastest-growing portion of the semiconductor market in terms of major segments. We have seen that trend over the last 20 years. We expect that trend to continue over the next decade for memory and storage to be the fastest growing. And as data becomes sort of the new oil of the new economy, I would say that, that sort of faster growth totally makes sense. And so we are in a very, very good position certainly in 2021. And even beyond that into 2022, we see this to be a set of secular growth drivers, not just a bounce back from COVID. Of course, there will be the bounce back happening from the weakness in the economy in 2020, but we are seeing strength across all of the segments you mentioned. Consumer is strong. We have a lot of investments happening on the data center side that relate to some of these new processors that are getting rolled out into the data center in 2021 that go from 6 channels to 8 channels of DRAM attach. So higher average capacities, a lot of strength in different parts of the market, the new console cycle on the graphics side. I mean, there's a lot of positive factors literally across the board, and it's tough to find a segment that is going to be -- expected to be weak. We have current weakness in the traditional enterprise space due to the economic weakness. But even that, we expect through the course of calendar 2021 will pick up pace as the economic recovery gathers momentum, and we should see broad-based strength across the board.

I appreciate that. Underlying that sort of high-level demand picture, your calendar '21 expectations of sort of high teens on the DRAM side, high teens percentage bit demand growth for the industry, underneath that, what's going to be driving -- maybe you can rank order what are going to be the growth drivers of that bit demand growth. Is it going to be -- there's servers, there's PCs, there's smartphone and auto and other markets. And then similarly for the NAND business where the team is expecting 30% type industry bit demand growth for this calendar year, what are going to be the drivers of that? Could be client SSD, enterprise SSD, mobile and so on. Maybe just kind of help rank order the drivers of bit demand growth in calendar '21.

Sure, Harlan. So I think, first, I just wanted to clarify that our view of calendar year 2021 has not changed in terms of demand growth, especially on the DRAM side, but the fact that we have called it out as high teens instead of 20% is an artifact of 2020, especially the second half coming in higher than expectations. So we had originally thought that it would be mid to high teens in 2020, and our latest view is that 2020 is going to end up above 20%. So the base has gone up from 2020, which has just artificially depressed the year-on-year numbers. But in terms of just the bit shipments, we expect, and the market demand for both DRAM and NAND, we don't have any changes. If anything, the bias is to the upside in terms of the revisions to the growth that are taking place. So I just wanted to start with that clarification. But then if you look at the segments that are growing faster than that average, certainly, the automotive segment is by far one of the fastest-growing segments. And again, it's a secular trend. You have a lot of driver-assist features rolling out, a lot of automation, a lot of infotainment improvements in the car. And certainly, as more EVs roll out, they have a dramatically higher semiconductor content, including memory and storage, compared to traditional cars. So that's clearly driving how new cars get designed, built, architected. So I think that trend is going to last a good part of the decade, if not more. So we have a view that the automotive segment will be the fastest-growing secular driver throughout this decade across memory and storage. I think the cloud market is another important market. We just spoke about some of the architectural changes taking place in the data center, a lot of new processors with high core counts, higher memory channels. So cloud is going to grow faster than the average, both on the DRAM and NAND side, in terms of SSD deployments on the NAND side. And we think that the mobile growth over the years is likely to be at or close to the average. Of course, we have in 2021 a special situation where we have a bounce back in the units from a decline in 2020. And then you have the multiplicative factor of higher average capacities as 5G ramps. We had originally said we feel we would do more than 200 million handsets in 2020. Looks like it's going to come in even higher than that and then more than doubling in 2021 in terms of 5G handsets, so 500 million or more. So that's going to be pretty robust. Now the segments growing slower than the average would be the PC segment and the traditional enterprise. Traditional enterprise, mainly because a lot of that growth is shifting towards the cloud. And the PC segment, again, mainly because average capacity growth there is slower and the unit volume growth is also relatively modest. So between them, we expect that those segments will be growing slower than the average.

Great. Perfect. And specifically on the mobile side, the Micron team has been talking about it for the last couple of quarters but didn't talk about it on the last earnings call, and that's Huawei, which was expected to be a headwind for the team through fiscal Q2. Should we just assume that Micron is past the impacts of the Huawei ban with other customers picking up the slack? And the other dynamic is Huawei did spin off its midrange Honor brand, new owner, non-Huawei. And the Micron team had a very strong design win pipeline with Honor. I assume that given Honor is no longer Huawei-owned that you are shipping your MCP solutions into these phones?

So yes, that's a good observation. We didn't talk much about Huawei. Just to clarify, again, I mean, we don't yet have the license to ship to Huawei on the mobile side. But since Honor is not Huawei, and we try not to comment on specific customers unless there are special situations like the one we had with Huawei, with major customers. But I would say that what you inferred earlier was right that we did have much better-than-expected progress in terms of placing our products with other smartphone companies, and the demand on the mobile side has been exceptionally strong and certainly much more so on the DRAM side in terms of inventory. Dave gave you some of those numbers, about 10 days reduction in inventory. But as Dave mentioned on the earnings call as well, we have changed how we do inventory valuation, et cetera. And so the number to think of from that 95 to 105 days of average inventory is really closer to 115 days of where we ended F Q1. So we don't have that much inventory to go in terms of getting to normal levels. And we have higher, somewhat higher level of inventory than that average in NAND and much lower than that average in DRAM. So on the DRAM side, things are very tight. We don't have any issue on the -- having the Huawei products be placed elsewhere. And in fact, in several parts of the DRAM segments of the market, we have shortages in several product categories. We are trying to maximize what we can to ship as much as we can because customers are really in shortages in certain areas. So I think on the NAND side, over the next few months, we'll get there as well, and we'll continue to see good improvements on the inventory front. So really rapid progress there and some emerging shortages happening on the DRAM side even right now.

And as a follow-up to that, I've got an incoming question from an investor. So the team talked about -- Dave mentioned this in his opening commentary. You guys talked about it on the earnings call. Team talked about seeing DRAM pricing increases in certain segments. Can the team articulate what specific areas they're seeing pricing increases?

Yes. I mean, we have strength in DRAM across multiple segments. We have strength in the leading edge nodes. We have strength in some of the legacy nodes, again, due to automotive strength on the legacy nodes. So there's fairly broad-based strength. And the dynamic that occurs is that, and the only one area that we have mentioned in the past that we have some near-term weakness is enterprise. But again, a lot of that slack has been taken up by demand from other segments. And so what tends to happen in this environment is, I mean, the market sort of tightens in this fashion and you end up with subnormal levels of inventory buffer, et cetera, then that pricing trend starts with a few segments but then spreads across the whole market. So I would hesitate to provide that as being localized to any segment. It's a typical start of a tightening phase where the shortages start to spread to multiple segments of the market.

So given the demand environment that you see ahead of you and we talked about for calendar '21, given the disciplined supply by you and your other DRAM competitors, you would expect probably a broadening out of some of the positive pricing improvements as we maybe move through the year? Is that a fair assumption?

Yes. I think it is fair to assume. But certainly, in our thinking of how we look at calendar year 2021 and the rest of fiscal 2021, and certainly, even looking beyond that, we expect that as the growth vectors continue to gather momentum and the economic growth continues to come in at a good clip and especially as the vaccine rollout starts to drive more economic activity, we certainly expect that the tightening in our markets will continue to gather momentum as well throughout the calendar year. So we do feel that because the supply has been, as you said, fairly disciplined on the CapEx side in DRAM, the slack in the DRAM market is very, very low right now, and we have a very robust growth expectation ahead on the demand side.

We just had the NVIDIA team present right before you guys, and they've got this entire new lineup of their 30 series gaming cards. They're ramping their data center products. The problem is they just can't ship enough. When you go to all the retail sites, you just can't buy a NVIDIA RTX 3070, 3080, 3090 or 3060, right? And we've seen multiple other semiconductor companies experiencing long lead times, shortages, not only in the graphics chain, but in the PC supply chain. In what end markets do you feel Micron's bit demand may be negatively impacted by the issue that your customers are not able to procure enough or other semiconductor products? And I'm not sure if you guys can quantify that potential impact or not.

Yes. I think that's a great question, and we are trying to ascertain some of that ourselves as well because things have changed quite rapidly over the last few months as we look ahead for 2021 calendar year. And I would say that there is no doubt that these shortages are starting to happen across various parts of the semiconductor supply chain. And from a memory and storage product perspective, which is what we build, certainly, for our F Q2 forecast, we don't feel we have any material exposure from a supply perspective for what we need to build for our customers. But of course, looking beyond that time frame, there are some shortages in the semiconductor supply chain that are concerning. And we are certainly working on it. Our supply chain team is working on it for all the parts that we need in terms of controllers and substrates and everything else, but also to make sure that we understand from our customers what are they seeing in terms of shortages. I mean, if you look at laptop and Chromebook shipments that are very, very strong, do they have adequate number of display drivers and other types of things that are in short supply, LCD panels, everything else. And so you've seen all of the things reported in the press on automotive shutdowns happening in terms of car lines going down because of lack of adequate supply. So there are certainly signs of shortages starting to creep up. So this year, I think a lot of companies will be chasing supply because demand is certainly outstripping supply.

On the NAND side, this is again coming from some investor questions. We've got a couple of them on this one. So on the NAND supply, the Micron team has talked about a little bit more supply discipline by the NAND industry competitors. Question is, does this take into account the Chinese competitors like YMTC? And one of the clients are saying that YMTC has targeted to ramp capacity to 100,000 wafer starts per month in 2021 and working on higher layer 3D NAND. And so is much of your sort of more guarded view on NAND and some of the supply discipline, is that more focused on the China competitors or just broadly speaking by all 6 or 7 of the other NAND competitors in the market?

Yes. I think when we think about the NAND market, and our comments mostly relate to the traditional suppliers who have been the 6 producers in the market. Because if you look at our bit growth, of course, in any given year because of how products ramp, bit growth may be more or less than the market demand growth. But over time, we have had our CAGR be very consistent with market demand. That's our goal. Our goal is to grow with the market in terms of our supply but grow our share of the profit pool through a lot of the activities around portfolio of products, mix, et cetera. And that's where QLC is very important, and we are making terrific progress there. I think in terms of how we look at the supply, it's really some of the other traditional players. In terms of China and YMTC, we have been following them closely and encompassing our views of what they will produce in our models. And so our expectations are always encompassing all of the potential supply that is likely to end up in the market. But that's not the part that is really the genesis behind the commentary on our supply. I mean, YMTC has all of these plans, of course, but they still have to prove themselves in terms of quality and reliability and being able to scale output and being able to create NAND at more than just consumer-grade quality. So there's a lot of work ahead on that front for them, but we have made some conservative assumptions. But I think as the industry supply growth moderates, I think there is an opportunity for NAND to improve as elasticity of demand kicks in on that front as well.

Great. And then maybe let's focus on the product and system-level focus of Micron. So we've been monitoring your share in the SSD segment of the market. The team has been gaining share in client and in enterprise. Enterprise has been a little bit more lumpy. But there, I think the customer buying patterns can be a little bit more lumpy as well. But in enterprise SSD specifically, we know that the Micron team comes from a position of strength with your SATA-based SSD products. And SATA is still a pretty big part of the overall SSD market, but the faster-growing part of the market is NVMe-based SSD. So where is Micron in terms of its qualifications on its next-generation NVMe products? Are you ramping them with customers now? And how should we think about Micron's share position in NVMe as we move through 2021?

Yes. Great question, Harlan. And we have mentioned in the past that we had gotten approximately to 80% of high-value solutions in NAND, and that was on the back of the strength in both SSDs as well as our mobile management portion of the business. So coming to SSDs, certainly, most of our client SSD business is NVMe, 90% of our shipments are NVMe. And we also mentioned that almost half of those are QLC now. So we are not only doing NVMe products in client. We have industry lead in terms of the portion of our client business that is QLC, which is very attractive for our customers and gives us better profitability as well. On the data center side, you're absolutely right. We have #1 share worldwide in SATA SSDs in the data center. NVMe data center SSDs is where a lot of our R&D effort is focused. We have existing products where we have gotten design win traction, and we continue to make progress there with customers. But we are really focused on our next-generation product, which is a vertically integrated product. We are going to introduce it over the next few quarters, and then we will be able to bring out products with future generations of SSDs built on the same architecture and that same base. So we will end up with much better efficiencies and that platform will mature over time. But Micron obviously was late to NVMe data center SSD with its own vertically integrated architecture. That was an effort that was put in place by the new management team just a couple of years ago. And so we obviously have to go through the development of that, which is a time-consuming thing, but we have confidence that we'll get to the other end of it and establish a much better business. And when that happens, we think the profitability profile of our NAND business will improve dramatically as we gain share in enterprise SSD and as QLC continues to become a bigger and bigger portion of our overall business.

A question for Dave on the financials. So structurally, it was a big initiative for Micron. The Micron team drove roughly $9 billion in structural profitability improvements fiscal '16 through fiscal '19. As we look out over the next few years, how should we think about structural profitability improvements in the Micron model beyond just taking down manufacturing and component costs? What are some of the other initiatives the team is doing to drive cost improvements across the company to drive margins higher?

Yes. Well, obviously, I mean, even though you said besides those, I mean, certainly, continuing to advance the process technology on both DRAM and NAND is incredibly important for us. It's our big focus. I'm really happy that we got to this leadership position on the leading edge with both DRAM and NAND really for the first time for Micron. So that's incredibly important. Sumit mentioned we're focused on the high-value part of the market, for sure. And generally, you participate in the high-value part of the market, customers pay more for that value. So that's still a very important part of our strategy. Also on the back end side, this is one that's maybe a little bit later in terms of our progression of it, but we have been working on bringing more of our back end in-house. Obviously, we'll always have some back end that's done through third parties but bring more of it in-house because we can get a better cost structure on the back end. That's all obviously also pretty important for us. Beyond just on the cost side or COGS side, we are also continuing to drive leverage. We'll probably from this year versus last year improve the OpEx leverage by somewhere on the order of 100, 150 basis points. So that's also important. We do want to see some leverage. Even though we are pretty effective and pretty frugal in terms of our OpEx spend, there's still some opportunities there. So I mean, this is a journey. We're never done with improving cost. We have now with the new replacement gate technology, now we have a very good cost structure. There's still some improvement to be made on the DRAM side relative to the industry, and we do want to continue to make improvements there.

That's great. Well, guys, we're just about out of time. Sumit, Dave, thank you for joining us here today. Best of luck in the new year. Looks like it's setting up to be a strong year for the Micron team. So thank you very much for participating, really appreciate it.

Thank you for having us, Harlan.

Yes. Thank you, Harlan. Take care.

Yes. Thank you.