All right, why don't we go ahead and get started? Good morning, and welcome to the first day of J.P. Morgan's 52nd Annual Technology and Media Communications Conference. My name is Harlan Sur. I'm the semiconductor and semiconductor capital equipment analyst for the firm. Very pleased to have Wissam Jabre, Chief Financial Officer of Western Digital, Peter Andrews, Vice President of Financial Planning and Investor Relations, also here with us today. I'm gonna turn it over to Peter, who's gonna read the safe harbor statement, and then we'll go ahead and kick off the Q&A. Gentlemen, thank you for joining us this morning. Peter?
Thanks, Harlan, and thank you for hosting us today. Before we begin, we'll be making forward-looking statements in today's discussion based on management's current assumptions and expectations. These forward-looking statements are subject to risks and uncertainties. Please refer to our most recent financial report on Form 10-K and/or other filings with the SEC for more information on the risks and uncertainties that could cause actual results to differ materially. We will also be making references to non-GAAP financial measures, and a reconciliation between the GAAP and non-GAAP financial measures are available on our website. So with that, let me turn it back to you.
G reat. Again, thank you for joining us this morning. So I'm gonna start off with t he near term, near to mid-term demand environment, t he cyclical dynamics of the sector. You know, after a sharp downturn in storage and memory, the industry is at the early stages of an upturn, right, in your HDD and flash businesses. Prior to that, the team had been under shipping market consumption for, I think, almost 8 consecutive quarters to the H2 of last year. After this 7, 8 quarters, demand is picking up, pricing is improving. How does the team see the business environment as we transition into the H2 of the year and move into calendar 2025?
Good morning, Harlan Sur. Happy to be here. Thanks for having us. So on the business side, we've continued to see improvement in the demand environment. When we look at the flash business, we last quarter, we're starting to see improvements in the enterprise SSD side, so cloud is coming back. And as we go into the H2, obviously, we have the consumer and client seasonality of the business that we expect to continue to see, in addition to continued improvement and the enterprise SSD on the cloud side, which could have a bit of upside for us. On the hard drive, the cloud demand has come back with a lot of deployments on the data center side.
What we saw in the March quarter is a return to normal revenue patterns, and we continue to anticipate improvements sequentially for the remainder part of the calendar year. So all in all, we feel good about the demand environment on both businesses.
The team issued a letter to customers at the beginning of April, outlining the strong demand increase, along with supply constraints and the resulting potential increase in pricing for both HDD and flash. Has the supply situation actually become tighter since you put out the letter? Some sources suggest that this supply tightness will persist into next year. You know, given the team's improved visibility and market demand trends, capacity, discipline, do you anticipate that demand could continue to exceed supply into next calendar year?
So, when we look at the hard drive side, I guess on both sides of the house, we're seeing tighter supply. So on the hard drive, in our hard drive business, we've taken action to resize our manufacturing footprint, and so that, basically, we resized it to where we see the near and medium-term demand going.
We've been working closely with our customers on build to order, which would allow us to get a bit more visibility and predictability, and also allows our customers to get more predictability in when they would get the products they need. On the flash side, in a way, similarly, we're seeing demand outstripping supply for the rest of the calendar year. And so there, too, we continue to be proactive on where we wanna place our bets so that we continue to optimize our profitability. But from a demand-supply perspective for our business, we see demand continuing to be higher than supply this year.
You know, AI and accelerated compute have been big drivers of the semiconductor industry recovery, right? It's created roughly $30 billion-$40 billion of GPU and AI ASIC semiconductor revenue demand that wasn't around two years ago. By the end of this year, for example, we forecast that 18%-20% of all industry DRAM wafer starts will be targeted for High Bandwidth Memory that goes into AI servers. Hard disk drive, you know, is used to store the massive data repositories used to train these complex AI models. We know that enterprise SSDs are used in the critical loop, like for both training and inferencing computations. Can you quantify the incremental demand pull from Gen AI on both nearline HDDs and also enterprise SSDs? I know, for example, some of your competitors-...
have said that, you know, there's gonna be 3x more NAND content in AI servers versus general purpose servers. Does the WD team have any similar metrics?
So, what we see is, we think AI will be impacting eventually various parts of the portfolio, be it on the consumer and client on the edge, as well as in the core, in the Enterprise SSD side. On Enterprise SSD, you know, last quarter, we saw the beginning of the return of the cloud. And as we get into the H2 of the year, we're potentially gonna see the beginning of the AI impact on Enterprise SSD from Enterprise SSD. Sorry, from AI workloads. So we have a few products that are already qualified, where we're seeing our customers coming back and asking for higher capacity points for those qualified Enterprise SSDs.
But we're also sampling the PCIe Gen 5 BiCS6, which is our next technology or our newest technology node in terms of Enterprise SSD, where we see potentially some of these products qualifying and starting to ramp in the H2 of the calendar year. So, from the training to inferencing, as well as storage and retention, when we think of the storage and retention, there's obviously that drives quite a bit of demand for the hard drive business. But at this stage, I would say, we don't see it yet in our results. What we're seeing is just a stronger demand from the normal type of business.
But we expect as we get more and more of the AI workloads becoming pervasive, we see a bias to the positive with respect to the demand picture. I know your question was mainly focused on servers, but also remember that AI is likely also gonna touch the cell phone t he PC, gaming, and a lot of the other end markets, where we also have a big exposure from our flash business.
That's right. You know, as I've talked to investors, and again, we're early days, right? And I think this is still... The use cases are still playing themselves out. But there is a little bit of concern at the margin around, will there be cannibalization, right, of Nearline, Cloud-Optimized HDD, in, you know, replacing, being replaced by Enterprise SSD? I can tell you that based on some of the observations that we've seen, the answer is no, right? Enterprise capacity optimized HDD, still the most optimal, most cost-effective solution for storing these huge data repositories of just the raw data, right? And enterprise SSD is situated a lot more closer to the compute clusters, right, and a part of that. And so is that how you guys see it as well? It's very complementary and not so much one cannibalizing the other.
T hat's how we see it, Harlan. We don't see cannibalization. Basically, we see different use cases more tied to whether it is Enterprise SSD or the hard drive. And so as we think of the demands, actually, we see a positive bias on both Enterprise SSD and hard drive.
And also, don't forget, for hard drives, you know, I know the question was, again, training and inference, but remember, there's just general collection of data.
That's right.
I think as we look into the future, more and more people are gonna store, store more data just in case in the future, they're gonna wanna use that to put into those training and inference engines. In addition, remember, with all these models, they're gonna produce something. They're gonna produce video, images, who knows what, that you're gonna wanna store. So I think even if you look at the two, two other ends of that training and inference, t here's great opportunities for HDDs.
As we look into the H2 of this calendar year, I know, Wissam, you said the team is confident on driving sequential growth in nearline or capacity HDD. Given your expanding lead times, better visibility, how are you seeing the potential exabyte growth looking into next year, right? Because I know the team has talked about normalized exabyte growth of around 25% on an annualized basis, but we're coming off of a low base. It feels like demand is strong. Could the 25% exabyte growth be conservative looking into next year? Just again, given the dynamics of coming off the bottom of a cycle, strong demand trends, could we see an overshoot to that 25% growth profile next year?
So, look, it's a bit too early to necessarily talk more specifics about next year, but where we see the exabyte on the, on the nearline side, where we, where we see the exabyte demand is it's in 20%-25% range. Obviously, we're gonna experience peaks and troughs through cycle, and so there'll be times where the demand will be higher than that, times where it'll be lower than that. But what we see from a long-term demand is the 20, or growth is 20%-25% range.
On the flash side, from an industry perspective, I know the team has been anticipating mid- to high-teens % increase in calendar 2024 bit demand growth and a modest, right, mid- to high-single-digits % increase in industry fab supply output, indicating continued strong supply and CapEx discipline by you and your peers. Given that, you know, we've got a little bit more visibility into the H2 of this year, is this how you still see industry supply and demand dynamics playing out for this year?
We still see a similar type of dynamics as you described them. And as we get into the H2 of the year, if I bring it to our business, we do have strong presence in consumer and client, and H2 of the year is a high season. It's a great, obviously, it's the seasonality of will drive that demand for us. And as I mentioned earlier, as the cloud demand comes back, we continue to anticipate upsides from the Enterprise SSD side, so we could potentially see higher numbers there.
Having said that, we've said a few times now that we continue to focus on being proactive on where we place our bets, because we do want to continue to improve our profitability and the profitability of our business. So there could be some variability there, but that's really high level, how we see the H2.
I feel like one of the big differentiators, because we've been covering WD for a while, we were covering SanDisk for a while as well, and I feel like the flash portfolio of the WD team is very differentiated, right, relative to your competitors. And what I mean by that is, the team still enjoys a very, very strong presence in client and retail, right? The SanDisk brand, still number 1 in the world, right, in retail. And I feel like that was an added venue for the team to deploy bits in an environment where you could selectively choose highest profitability pools, right, as you manage through this downturn. I do think that SanDisk, I do think that WD does have the highest client and retail exposure relative to all of your NAND flash peers.
But is that the case? Is that considered to be a big differentiator for the WD team, having that client and retail portfolio?
I mean, we see. We do have good presence on the flash side and consumer and client. And yes, the SanDisk brand is a premium brand that is well recognized when it comes to flash. And so that has actually helped us quite a bit, as we were weathering the deep cyclical downturn. But we continue to also create good differentiation within that product portfolio to enable us to continue to have strong presence in consumer and client.
As it relates to, you know, the team has been very disciplined. You've been maintaining flattish, sequential bit shipments now, to the last few quarters, including this quarter, right? Can the team actually expand bit shipments in the H2 of the year to align with what looks to be, as you mentioned, seasonally stronger H2 of the year, and just an overall, you know, better enterprise demand pool? Can the team expand bit supply in the H2 to match that demand curve?
When you look at our supply, we've built a little bit of inventory exiting March, and we also anticipate a bit of inventory build in the June quarter. But that's all to get us prepared for the H2. We're comfortable with the supply with our supply at the current point, given the environment, and where we see the H2 demand coming. So there's no... I don't see any issues with us being able to supply the products needed, as well as obviously, deliver to the demand.
On profitability, you know, the team has maintained very strong focus on two cycle profitability in response to this downturn, right? Implemented some structural actions, right, like on your HDD franchise, and some temporal actions, like modulating utilizations, as you mentioned, reallocating bits to higher profitability applications and segments. Take us through the structural and temporal actions and, you know, how they've benefited the WD team as industry fundamentals now are also starting to improve.
So, on the hard drive side, which structurally reduced our manufacturing footprint, which basically helped us reduce our fixed cost. And so as the business and as the demand and the cloud demand comes back, we anticipate to generate better profitability at lower volumes. And actually, as we've demonstrated in the March quarter, our hard drive business delivered 31% gross margin.
So which is already at the lower end of our target range, but we've gotten into the target range. And when you think of it, it's really at a lower revenue than we've delivered similar types of gross margins in the past. In the past, the 31% gross margin would have required much bigger top line revenue. So that was helped by some of the restructuring of our manufacturing footprint. On the flash side, throughout the cyclical downturn, we reduced our output to limit the inventory build, and also obviously conserve cash. And as we start seeing demand coming back, that is going away.
From that perspective, there I would say the demand for our products relative to where the supply is helped with the pricing, helped improve our gross margins. And then when you look at the total company, we've also taken quite a bit of action to resize our operating expenses, and so that also will help position us in a much better place as we see demand come back, revenue come back, gross margins being at a higher point. And so we should start seeing our operating margins also benefiting from that.
I mean, the other key thing in there is the product portfolio.
Right.
Especially in the HDD domain. I mean, if you look there, we've got the best product in the market, you know, as the cloud is transitioning over to SMR drives. We've got the highest capacity, highest volume, and the best cost structure, and you're seeing that flow through, especially in our HDD results.
Considering that, following up on that, you know, that you just, like you said, drove within your through-cycle HDD gross margin target range of 31 to 40 -- 34% last quarter, both with and without underutilization charges. You know, foreseeing continued growth in exabyte unit shipments, continued cost downs, how should we think about your gross margin trajectory exiting this year into next year? And with the structural adjustments that you just talked about, right, it seems that the team can see a through-cycle gross margin profile that is actually higher than your current through-cycle range of 31%-34%. I'm just wondering if the team can also address that as well.
So, you know, March quarter, we've—as I mentioned earlier, we've gotten into the target range. We're not at a point where we're. I mean, I'm not here to update our t arget model, on the fly. But we continue to focus on, on the profitability of the business. And as Peter mentioned, obviously, we have a great, portfolio on the hard drive side. You know, we continue to, commercialize our, our roadmap in ePMR, OptiNAND, UltraSMR, which really provides, the best TCO and best capacity, and the highest capacity points for our customers. And as we continue to execute on that, that will, that will - I mean, this is a technology that, we've been manufacturing for many, many years. It's well understood. It's, high quality. It's, very reliable, and so, that will help us continue to drive better profitability as well.
Also, those margins are through-cycle margins?
That's right.
Remember, we've operated below that range for quite a while. Hopefully, we'll have some quarters where we'll be above it, average out somewhere in that 31%-34% gross margin range. Maybe, maybe then more near term, as we think about the H2, your commentary about continued exabyte shipment and unit growth through the second, the calendar H2 of this year. The team is obviously driving, continuing to drive cost downs. Can we expect a positive bias on gross margins on HDD as we move through the H2 calendar of this year?
W ith the continuing dynamics that you just described, t here's no reason for us not to expect a positive bias on the gross margin.
On the flash side, you know, pricing continues to improve, continue to drive mid-teens% cost downs. You're within striking distance of moving into your flash through-cycle gross margin range of 35%-37%. What's the trigger point, profitability, ROIC metrics, or visibility that you need in order for you and your JV partner to start to increase capacity and maybe spend a little bit more CapEx?
Today, we're focused on the profitability of our business. We've gone through a few quarters of really deep downturn, and we're basically looking at continuing to improve the profitability of the business. As we said before, we'd like to see our business execute to the through-cycle margins of 35%-37%. And so, as Peter just mentioned, you know, this is really a through-cycle number, which means, you know, we need not only to get into that range, but we need to be higher, above that range for quite some time to be able to see that 35%-37%. That's really, that's really where what we're focused on short term, with respect to the flash business.
Before I start talking about, I wanna get into some of the product and technology areas. But before that, I wanna open it up to the audience to see if there are any questions. If you do have any questions, just raise your hand, and we'll get a mic over to you. Any questions? Oh, we've got one. Okay. Can you talk a little...
Can you hear me?
Yeah.
Can you talk a little bit about what you expect, the industry's capital deployment plans, to look like over the next, I don't know, 18 months, 2 years? This is a very good time, and it seems like storage companies are tempted to spend a lot on increasing capacity in this type of environment. Is this time different?
Is your question related to? Ri ght, on the flash side.
CapEx on both sides. But mostly on flash and just not just for Western Digital, but for the industry.
I can't comment on the industry. I can comment on how we think about CapEx for both flash and the HDD business. So for the flash business, it's really, for us, the focus is to continue to improve the profitability of our business before we're ready to increase the investment in capital. We've gone through several quarters of really negative numbers, and so we're... We need to see that gross margin for the flash business be within the through cycle 35%-37% before we're comfortable to make additional capital investments or incremental capital investments. On the hard drive side, I would say same dynamics for us.
You know, we think we have the appropriate manufacturing footprint for the near and medium term, and we continue to focus on improving the profitability of the business. We have a great portfolio for both businesses, whether on the hard drive with the highest capacity points in the industry, as well as on the flash side, we have probably one of the strongest portfolios in that business for in many, many years. And so there's no reason for us to expect to continue to see improving profitability before we get into CapEx.
Just a few additional data points on that. Remember, in FY 2023, we took our total cash CapEx down over 30%. FY 2024, we're on track to bring that down another 50% off that 33%. So we've taken dramatic actions to pull back our CapEx in general. Again, the key focus is profitability.
I think, do we have another question around here? I think there. Perfect.
Thanks for the opportunity. I just wonder, can you comment on the JV with Kioxia? You know, how is the relationship so far? And I guess just trying to understand, you know, what are the options in front of you for that JV business in terms of... I think there has been some rumors about maybe combine the business with the other partner, but if you can share anything about that. Thanks.
I can't comment on rumors and speculations. What I can talk about is our relationship with our JV partner. The joint venture with Kioxia is important for the flash business. You know, between the two of us, we develop memory technology together. We have economies of scale when it comes to the manufacturing side, so it is really an important relationship, and it's been a great relationship throughout the years. It continues to be strong, and we continue to operate in really good partnership.
The key thing for us is to focus on the spin where we are today. That's still on track for the H2 of this calendar year. That's where our eyes are focused right now.
Of course.
Any other questions? Let's transition over to the product side. You know, we continue to get questions on the team's transition to HAMR-based HDD technology. You know, with some concerns that you're behind, right? WD's move to HAMR, I feel like, is similar to SanDisk's. If you remember back in the 2013, 2014 timeframe, SanDisk's slower move to 3D NAND, because SanDisk could extend the life of their 2D NAND well beyond competitors. And in hindsight, it actually turned out to be the right move for the SanDisk team, as NAND competitors all struggled with bringing 3D NAND to the market in terms of yields, in terms of manufacturability.
In a similar fashion today, WD is able to drive high 30s terabyte-type capacities with its SMR, UltraSMR, ePMR technology, and drive competitive TCOs for customers, and therefore, push out the transition to HAMR, which allows the WD team more time to optimize HAMR from a manufacturability, cost perspective. Is this reflective of how you think about the strategy on HAMR?
So when, when it comes to HAMR in the HDD business, we, we view this as a technology decision as opposed to a business and portfolio decision. We've continued to execute on our roadmap and our portfolio, our portfolio strategy. You know, as we've demonstrated with ePMR, OptiNAND, UltraSMR, we've continued to, to increase those capacity points. We do have line of sight to the 40 TB as we continue to execute on this strategy. You know, as the last quarter showed, obviously, with this type of portfolio, we can generate great profitability for our business.
So having said that, you know, our team also continues to work on HAMR development. They understand HAMR very well. They understand what it takes to qualify. It is something that we work very closely with our customers on. But we see HAMR a 40 TB, a 40 TB-plus story. You know, our focus is on really the right products with the right cost at the right time.
Yes.
So we want to be able to continue to provide the best products at the capacity points and provide a good TCO for our customers. And, you know, as we've shown, that is very well received by our customers, and they continue to obviously buy, and our portfolio continues to perform really well. So in the context, as I said, of the right product, right cost, and the right time, this is where we see the 40 TB is would be a good intersection for the HAMR products.
With the ramp of 26- and 20-terabyte UltraSMR this year, where are you sampling in qual with in terms of capacity points for drives and solutions that your customers want next year? Maybe a little bit of a snapshot into the, the roadmaps over the next 12 months.
So when you look at the 20+, we have, we have today 4 drives in that 20+ TB category. This is pretty much now making up most of our nearline shipments. And with UltraSMR, obviously, it does provide a great TCO for our customers. In fact, when you look at the SMR, we, last quarter, almost 50% of our nearline business was or enterprise capacity business was in the UltraSMR. So we continue to make good progress with customers. Our customers, we have several customers that are sampling the product, and obviously will be ramping into course.
On enterprise SSD at its peak, right? This is a $20 billion market opportunity. WD consistently captured, about, you know, mid-single digit enterprise SSD market share. You talked about on the last earnings call, seeing demand pick up at the beginning of this year, partially driven by inventory clearance, cloud spending pick up, AI demand. You grew ESSD in March, but off of a low base. Several of your ESSD cloud programs you had prior to the down cycle are back in qual, higher densities. You expect to ship these in the H2 of this calendar year. You also announced that you're in qual with a hyperscale customer to your next gen PCIe Gen 5 platform. So seems like you have a tailwind here on enterprise SSD. What did the team do during the downturn to improve its competitiveness in enterprise SSD?
Can the team break out of its historical share range and make progress towards your long-term goals of mid to high teens plus market share, that you outlined at your last Investor Day?
So, you know, as Peter mentioned, obviously, we have a great portfolio, and we continued throughout the downturn to focus our—as much as we've reduced our spend, we continued to focus on areas that are important for our business and to basically strengthen our position and get us in a position to be stronger as we've started to experience the cyclical recovery. On the enterprise SSD side, you know, we do have already a few, some few qualifications with the, with cloud customers. And, as I mentioned earlier, they're coming back to, to, to ask for higher capacity points on these qualifications. We are sampling PCIe Gen 5 in BiCS6.
Yes.
So that's the next, the next level of the enterprise SSD portfolio. And as we get into the second calendar year, we expect to to start ramping some of these products. And so, from a market perspective, we started to experience an uptick in the cloud in the March quarter. We continue to expect strength in the next few quarters, and so, potential good upside in the H2. Having said that, obviously, our goal isn't necessarily to focus on a specific market share.
Yes.
Our goal is to optimize the profitability of the flash business. And with that, that means we will continue to be proactive and dynamic in placing our bets, which is really our portfolio strategy in terms of executing the quarters. And so, if the Enterprise SSD business has the good profits, obviously this is the—we'll will be shipping more bits there as well.
Your 3-5-year financial targets that you put out at Investor Day, back in 2022, right, 7%-9% revenue CAGR, 33%-36% gross margins, 17%-22% operating margins, was obviously before we entered the downturn in 2023. Did the team factor in a downturn over the 3-5-year horizon when you set your targets? And do you still see the targets as attainable by the 2026-2027 timeframe?
Those targets are very much through cycle targets, and so that basically factors in downturn as well as cyclical recovery. You know, having said that, obviously, we are working through our separation, and it's important for us to focus on each of the business. As we get closer to that separation and spin-off date, we will be coming back with with a clear view of where we see the targets for each of the businesses. But for us today, the targets are are still good. We're executing to them. We think they're they're through cycle, and we continue to basically make progress towards that.
Well, we are just about out of time with some. Peter, thank you for your participation today. We look forward to monitoring the progress of the team this year, and keep up the great execution. Thank you.
Thank you so much for having us.