Good morning, everyone. I'm Tom O'Malley, Semi and Semi-Cap Analyst here at Barclays. I just want to welcome everyone to the Barclays Tech Conference here. First, on the day, Western Digital. We've got David Goeckeler, CEO, Wissam Jabre, CFO. Thank you both so much for being here.
Great to be here. Thanks for having us.
Yeah.
So, I think that the way that I would like to start is just you've had this period of a strategic review. You guys have been mum for some time about what went on during that process. Could you just, one, talk about what was assessed? Two, why things ended like they did, and just any additional comments since you announced the split of the two businesses?
Sure.
Sorry.
I'd be happy to do that, but-
Yes.
Before we start-
Disclosure, I might say-
Wissam has some very important-
Disclosure
Information for us.
As advised by my legal team, we'll be making forward-looking statements based on current assumptions and expectations, and I ask you to refer to our most recent annual report on Form 10-K and other filings with the SEC for more information on risks and uncertainties that could cause actual results to differ materially. We will also be making reference to non-GAAP financials, and the reconciliation of our GAAP and non-GAAP results can be found on our website. Thank you.
Perfect.
All right, the strategic review.
Yes.
Yes. Very thorough process, a very detailed process. It was great to go through it. A lot of people participated. I will say that all the details inside of it are still under NDA, but I'll give you some more color around it. You know, we've been driving an enormous amount of improvement in both of our franchises over the last several years, just in the execution of them, the way we're organized internally. We focused on our balance sheet. We've cleaned up some long-standing problems that were kind of overhangs on the business, and I think got ourselves in a very good position to really figure out how can we unlock the value of these franchises? Because we think they're very, very valuable franchises. I think, quite frankly, the downturn showed that both portfolios.
I'm sure we'll get into it a little bit more. Both businesses are performing best-in-class. In the NAND business, significantly so, and even in the HDD business, where we've opened up a profitability gap with our peer companies there as well. So when you looked at all the different options that we had, what we were able to execute, given, you know, the environment we're in, we decided that separating the businesses was the best alternative for realizing value, and there's a number of things behind that. One, the execution of that plan is completely within our control. There's no regulatory approval required, any of that. Two, it unlocks value, what we think are tremendous franchises. Three, like I said, we've done a lot of work to restructure the businesses and get them in great shape and, like, get them ready to stand alone, on their own.
And then we've got a recovering market in both businesses. So when you put all that together, we thought it was the right time to make this move on the portfolio, and we're very happy to give an announcement, and on to the execution.
Great. Let's just start on the flash side. So could you just start by walking us through end demand in your key end markets, so client, consumer, cloud? And then talk maybe a little bit about the inventory positions in each of those markets. I think that'd be a good place to start.
Sure. So first of all, I'll say, you know, the bottom's behind us in NAND, and the recovery is on. I think the question at this point is just how fast. And so we'll get into that and what the dynamics of that are. But if you look at our markets, again, I think one of the big strategic advantages of our NAND business and why did we perform so well in the downturn and continue to perform so well, is just the diversity of our portfolio and the number of end markets that we can access. So start with consumer. That was kind of the first in, first out. Consumer market is back to, you know, what we would say, just regular normal behavior. Inventory levels look fine across all the different channels and all the different retailers we deal with.
We have hundreds of thousands of points of presence for our portfolio around the world, so it's a very big, diverse business. We're in a seasonally strong quarter, so we feel very good about that market. Inside that, we have a lot of great brands and a lot of work we've done on that over the last several years are all performing well. SanDisk, SanDisk Professional, our Black brand on gaming, just a very, very. That business is performing as expected, so kind of past the downturn and kind of leading us out of it, quite frankly. The PC client market, again, back to what we would think normal behavior, the way that customers are buying. You've seen customers in the last quarter return to more normal inventory positions. A lot of that has been, people have talked about that as strategic buys, things like that.
But really, for us, it's just kind of returning back to carrying a typical amount of inventory to run my business. We came through a very deep downturn where, you know, a lot of our customers went into that with too much inventory. They corrected quickly. They waited to see the return, the business come back, and we're back to that now. So that business, we feel good about where that is. You know, record Exabyte shipments. And then in the cloud, in the NAND business, we're still in an inventory digestion phase.
That business has not really come back yet. I think that's what's been one of the surprising things of the downturn in the NAND business. I think going into the downturn, everybody thought enterprise SSD is the business you have to be in. It's the, it's the most you know, the most important business in NAND, and it certainly is a great business. We have great products there, but it is the business that suffered the most in the downturn, and it's still... We're waiting for that to return to normal, and we think there's still some inventory to work through there over the next couple of quarters. HDD, you didn't ask about HDD, but we're starting to see the data center market in HDD start to normalize.
Great. Yeah, just, just diving in just a little bit there on the NAND side, in terms of the cloud recovery, do you have an expectation for a timeline? You said a couple of quarters, but is that when inventory gets to a better position, or do you think that you start seeing ordering patterns get a little bit better?
I think we'll start to see ordering patterns. From the information we have, you know, it's still a little early, but as we move through 2024, we expect things to get better.
Got it. So two of the three end markets seem good. The third is a couple of quarters away, but that obviously plays into the ASP side. If you look historically, when you look at a recovery, you get high single-digit sequentials for only a couple quarters in terms of the recovery and when- what pricing moves off of that recovery. If, if you look today, some peers are talking about 20%+ ASP increases. How realistic is that? How long do you think a pricing recovery can sustain in this example, just given how bad this downturn was?
Look, I mean, I don't think there's a playbook for the recovery. I mean, when we look at that this business, when we look at it from the inside, when we talk to our sales teams and the directions we're giving everybody, we had a downturn that was pretty unprecedented. So I think going back and trying to map other recoveries on top of this is really not the right way to look at it. From peak to trough, we had pricing down over 50%, 50%-60%, and that peak was kind of a mid-cycle number in the industry. So, you know, if you're down 50%, you need to double to get back to where you are.
So, you know, I think we need to see pricing, depending on the market, 'cause every market went down different amounts. We need to see pricing come back 50%-100% before you're gonna see capital flow back into this industry. So how fast is that gonna happen? We don't know. I mean, it could happen very quickly, depending on how the, you know, the market is. Bit supply is significantly under bit demand. You know, we've seen an acceleration of shipments over the last couple of quarters, which is obviously what you're gonna see in the beginning of a recovery. So I don't think we have a playbook on that. So how possible is it to see numbers that are bigger than in the past? I think it's very possible. Is it gonna happen?
I think we're gonna see over the next couple of quarters.
Yeah, I think it's a perfect time to jump in just to the supply-demand conversation. So what do you think bit demand looks like into next year? What do you think the industry will need to do to match that on the supply side? Are you fully intending to match that? I think some have talked about potentially, you know, a time period where digestion needs to take place. How long do you think you need to be below industry demand in terms of getting the market back to a healthy spot?
Well, we, y ou know, last quarter, we've upped our demand for this quarter. We look at calendar year 2023, you know, we were in high single digits, we're now mid-double-digit range, like, low to mid-double digit on demand side, given the significant acceleration in shipments in the last couple of quarters. Supply, fab-out supply, we see as negative single digits-
Okay.
In calendar year 2023. Calendar year 2024, we see demand high double, you know, high teens. We see fab-out supply still mid-single digit kind of number. So, you know, I think, you know, I'll speak for us. I think when everybody looks at what's happened in the downturn, there's an enormous amount of capital invested in this industry. The industry has lost an enormous amount of money in this downturn, so I think we're gonna look at this very, very carefully about when we put capital back in this business to meet the bit demand. It's gonna have to be more than, "This is just as what we project." We're gonna have to see these, we're gonna have to see the recovery come back, and we have to understand what that looks like.
Obviously, as we, what we talked about earlier, enterprise SSD and the cloud market coming back into NAND. It's a big market. It's an important market. We have to see how that develops over the next several quarters. And then, you know, I think the industry is gonna work very, very hard to get supply, demand balance correct and not get ahead, not get ahead of the game like we were going into this downturn.
So with that gap of production and demand, what does that mean from a utilization perspective? At what point do you start to increase utilization? What are the metrics that you're looking at that would make you get more aggressive? And then, could you make any comments on what you think about CapEx in the flash business for next year?
Yes. So, on utilization, as you know, in January, we announced a 30% reduction in supply, in wafer inputs, and since then, we've been managing our utilization in a very dynamic way. We look at the demand of our products, we look at what inventory is doing, and of course, we're focused on preserving cash. And so these are really the key things we look at as we look to continuously, dynamically manage our utilization. Where we are today, it looks like we'll continue basically the same process. As we see demand for our products improve, that would influence the decision on what we do from a wafer input.
How do you see the introduction of BiCS8 impacting the CapEx side over the next couple of years? And then, what's the timeline, that represents a more substantial portion of your bits?
So on the CapEx, which was also part of your previous question, on the CapEx in fiscal 2023, we did reduce our CapEx by more than 30% relative to the previous year. In this fiscal year, we're reducing our cash CapEx spend significantly by probably 50% or more. And as we see the recovery, let's say, taking hold, then we will be adjusting. But with respect to BiCS8, BiCS8 will probably be... It is not, it's pre- it's basically our next big node after BiCS6, and from a cash CapEx or a CapEx perspective, I would expect it to start.
Again, if the recovery takes hold and we see the profitability, then we should see CapEx for BiCS8 at some point in the H2 of fiscal 2025.
Got it. Got it. So we've walked through all the moving pieces, spinning out on the other side, you have this positive gross margin guide in the December quarter. If you look at other memory players, they seem actually a little bit behind of the curve that you guys are on. Could you maybe dive in and explain why you're seeing the acceleration in gross margins that you are? Is that just related to the JV, or how have you been able to come up the curve a little bit faster during this recovery?
Yeah, it's the result of a lot of hard work over the last three or four years, and changes we've made, and then a lot of hard work over decades with the JV. So let me decompose it a little bit. In the NAND market, just big picture, there's two things you gotta get right. You gotta get your fundamental technology correct, and you gotta get it in the right position, and you gotta get it in the right amount of CapEx to build that technology. So that's the BiCS roadmap, and that's a place where we invest jointly with Kioxia. It's one R&D team. So basically, we punch above our weight. We're the largest. You know, we have the largest investment. You know, just from a bit share point of view, we're able to invest like the largest player in the market.
That gives us a very, very good base technology position if you look at the BiCS roadmap. If you look at CapEx per output of bit, we're consistently, over the last five years, a third lower than the industry average. So we've got a very good technology base to build on top of that's very capital efficient and very high-quality product. Now, the other side of that work we've been doing over the last three years is you've got to figure out what portfolio. You know, I'm gonna take these wafers out of the fab, and I'm gonna sell them to somebody, and you can. There's lots of choices you can make. You can just sell the wafers, you can sell components, you can build controllers and sell SSDs.
So we put an enormous amount of work into what is the right portfolio to get the best outcome, and we're able to... The consumer portfolio that you said earlier is a real gem of the portfolio, so better through-cycle margin than any other part of the portfolio. Again, we have tremendous brands there, a lot of work on those brands over the last three years. So that's a great place for us to mix into. Client SSD, we have the second-largest client SSD portfolio in the industry, so a very, very good position there. We sell into mobile. We have a very good position in gaming. The WD Black brand, again, both into the consoles and retail, directly to the consumer.
That, you know, when I came to the company, the amount of bits we sold into mobile was an easy number to remember. It was zero, and now it's double digits, so that's a big part of the portfolio. And then we're qualified in enterprise SSD. Now, that market is very depressed, but we still have the products there. So we've done an enormous amount of work that, on top of that base technology that's very capital efficient, very high quality, we put a very robust, diverse portfolio, and then we put the agility in the organization to change mix dynamically to get the best outcome. Because every market in NAND is a different size, different growth rate, different profitability, and those change throughout the cycle, so you have to constantly be mixing across this. And, you know, we're able to do that.
We put the agility in the organization to be able to do that very, very quickly, and you're seeing the results, right? We're leading... You know, we bottomed out in gross margin, you know, significant double digits ahead of anybody else in the industry, you know, 10, 20, 30 points ahead, and now we're the first one coming out into positive gross margin territory. So, again, this goes back to the strategic review and why we think this business is ready to stand on its own. It's the best NAND business in the industry, and we're demonstrating it every quarter with the results we're delivering.
Helpful. So why don't we switch over to the HDD side of the house? You mentioned earlier that you do expect a cloud recovery on the HDD side this coming quarter. So one, could you just talk about how that trajectory has changed since you talked about the recovery at earnings? And then maybe any color in nearline between the cloud players and others. Is there a different pace of recovery? Are inventory positions different at either one of those? Just help, you know, shape what is happening in nearline right now.
So, when cloud went into the downturn, there was just clearly way too much inventory that had been built up, so clearly there was some buy-ahead that was going on, you know, when everybody was worried about the supply chain, the pandemic, and all those issues. And then we went into, you know, a multi-quarter period where some of the biggest customers in the market just stopped buying. We have customers that would buy enormous amounts, you know, $hundreds of millions worth of product just go to zero. So it was all through working off that inventory correction. So we talked about this over the last couple of earnings calls. We expected sequential growth as we moved throughout 2024 as those customers came back into the market, and the inventory positions and drives really started to normalize. That process is on.
We're not all the way there yet. Again, the bottom of the HDD business is behind us. We expect sequential growth in Nearline as we go throughout the fiscal year, and, you know, the conversations with customers are very good on what that looks like. And, you know, the industry is going more to a build-to-order process. You know, we've taken a significant amount of infrastructure out of this business. You know, we were overinvested again in this infrastructure, especially, you know, we're kind of at the end of this client-to-cloud transition, and now we've got the footprint sized for what we need. Our customers understand that.
We're getting more visibility into what their demand looks like, and we have a lot of confidence that the HDD business is gonna grow from here, and we'll continue to see the margins improve as well.
How about the client and consumer side? I think we haven't touched on those yet for HDD. You mentioned on the last earnings call, I think at least consumer was stabilizing. How are those trajectories as we go through the quarter here?
You know, they're fine. I mean, they're... It's a decreasing part of the business. I mean, consumer, we're in, again, a seasonally strong quarter, so we'll expect consumer to do well. The client business... Again, during the pandemic, I think we had a false signal that everything that was a client, you know, everybody wanted a PC at the start of the pandemic. So we had this kind of false signal where client SSDs went up.... And if you draw a trend line from pre-pandemic to now, actually, the client HDD penetration in the PC is now below where it was coming, that trend line coming into the pandemic. So this business is transitioning to cloud. You know, there's still important markets.
There's still millions of drives sold into those markets every quarter, but again, we've taken 40% of our capacity out of client, and we're getting that sized right to meet that market. But again, like the client SSD market, more predictable past the inventory corrections.
Great. So just zooming out a little more broadly, again, I just wanna talk about the broad strategy in your Ultra SMR drives. You know, how are you seeing customer interaction at those higher capacity drives when your competitor is moving to HAMR, and are you seeing better traction at the high end? Is that working? Can you see the result of that strategy yet, or is that still kind of on the come?
No, it's working. It's fantastic, quite, quite frankly. I mean, if you look at the 26-terabyte drive that we launched two quarters ago, you know, last quarter, it was nearly half of our exabyte shipped, right? So that tells you that the market at the very high end is moving to SMR. Now, SMR is a technology that's been around for quite some time. It's just never been adopted at the very high end of the market because there's work to do on the host side to adopt it. And the really key in driving the adoption of that was our introduction of UltraSMR. So a traditional SMR technology gives you an extra 10% of capacity on a drive, and it doesn't matter what kind of drive. It can be a CMR drive, a PMR, ePMR, HAMR, it doesn't matter.
SMR is a different layer on top of that. So when we introduced UltraSMR a couple generations ago, now it's +20%, and you're deploying a 22 TB drive, you know, going to 24 TB drive. An extra 20% is quite a bit. So the hyperscale vendors started adopting this technology, and, you know, we've been saying for quite some time that this was gonna be the next leg of growth in the data center, and that's turning out to be correct. I mean, we'll get to HAMR. HAMR will come, but once we get to HAMR, SMR is gonna work on top of HAMR just the way it works now.
So, we're very, very happy with the technology and the strategy, and we've got line of sight with our current technology portfolio of ePMR, UltraSMR, for, you know, many, many more generations of drives that are cost-controlled, high yields, and meet our customers' requirements, quick qualifications. So we feel like the portfolio is in a great, great shape, and it's again, that's another thing, showing up every quarter in the numbers, right? Last five quarters, we've had outsized profitability versus our peers, and that gap is increasing.
So you mentioned HAMR, and your competitor talks about 1 million units in the H1 of calendar year 2024 for HAMR drives. How are you progressing on your roadmap, and how do you see the technological landscape changing? Where are you in your progression in HAMR as well?
HAMR is a technology. I think HAMR... You know, when I was in Japan a couple months ago, I talked to the gentleman that actually wrote the seminal paper on HAMR. I think it was 2002. So HAMR's been in development little bit, and this is the story of the HDD industry. I mean, these drives, the technology in this is incredibly sophisticated and can take decades to bring to market. And I think where we're at right now is we're, like, in that end game of when does HAMR, the commercialized technology for the data center, and it's gonna happen sometime over the next several years. Our technology is at a... You know, we've been working on it for several decades. When we need to bring it into the portfolio, we'll fold it in as a technology.
But as I said earlier, we have ePMR, right? This is, I think, something that maybe we haven't talked enough about. We went to ePMR, Energy Assist on PMR, several generations ago. We've now shipped over 600 exabytes of ePMR technology into the market. So it's really this ePMR plus Ultra SMR is giving us this roadmap that it gives us the flexibility to bring HAMR into the portfolio when we can get the yields, when we can produce it at scale. Look, let's look at the nearline market. The nearline market, the depressed number in calendar 2023, calendar year 2023, is somewhere around 40 million drives. The year before that, it was 60 million drives. So when you talk about 1 million drives, you're talking about a very, very small percent of this market.
We have a strategy where we can satisfy the entire market at these very high capacity points with technology that is already been commercialized, can be produced at scale in a very cost-controlled way, and that's why we're getting the increased profitability across our portfolio. When HAMR can be produced as a part of that portfolio, we'll fold it in, that we get the right economics.
Helpful. I think we have time for one more here, but just... You've already talked about most of those end markets on the HDD side, actually getting a bit better. Some of the smaller legacy businesses, obviously, working through periods of downturn still. But can we talk about what it means for utilization in the H2 of the fiscal year? Can you give us an update on how that's trending? And then on the HDD side, just given those utilization changes, when do you think you can get back to what you've stated as q- kind of a normal margin profile of that above 30% range?
So maybe I'll start with the second part of the question. When we normalize... Like, for instance, when you look at the September quarter and we normalize for the or adjust out the utilization charges, our margin was pretty close to around 30%. Similar to the flash side, we manage also our utilization on the hard drive side to basically match our the supply of our products with where we see the demand. Over the next couple of quarters, we continue to see underutilization, but I would expect it to be slightly lower than what we're seeing now. So as the demand continues to sort of improve, I expect eventually utilization will be phased out.
But for the rest of the fiscal year 2024, I still expect to see some underutilization charges.
Very helpful. I think that's all we have in terms of time. Thank you both so much for being here. Have a lovely rest of the week.
Thank you. Thank you very much.
Thank you.
Appreciate it.