Welcome to Citi's 2024 Global TMT Conference. My name is Asiya Merchant. I cover the tech hardware, tech supply chain here at Citi Research. We're very pleased to have David Goeckeler and Wissam Jabre from Western Digital. This session is for Citi's clients only. I'm gonna first turn it over to the management here to have just some quick cover statements, and then we're gonna jump into Q&A. Welcome, everyone. Thank you.
Thank you, Asiya. Good morning, everyone. We'll be making forward-looking statements in today's discussion based on management's current assumptions and expectations, including with respect to our product portfolio, business, business plans and performance, market trends and dynamics, and future financial results. These forward-looking statements are subject to risks and uncertainties. Please refer to our most recent financial report on Form 10-K and our other filings with the SEC for more information on the risks and uncertainties that could cause actual results to differ materially from expectations. We'll also be making references to non-GAAP financials, and the reconciliation of our GAAP and non-GAAP results can be found on our website.
Great. Thank you. Thank you, David. Thank you, Wissam. So I'm gonna jump right into it. I think top of mind is the spin that you guys are, have been talking to investors about. Maybe for those in the room, just a quick recap on the strategic rationale for spinning out the two businesses, and what should investors focus on as you look at these businesses individually?
So first of all, thank you for having us. It's great to be here. Well, you know, we spent a lot of time thinking this through, and, you know, I've been with the company a little over four years now and trying, you know, really focused on getting this business executing extremely well and unlocking the value in the company. And, you know, I think on that first step, we've spent a long time building the portfolios, getting the businesses in great shape. I think they're executing extremely well. You know, our HDD business is the largest, fastest-growing, most profitable HDD business in the industry. Our portfolio is in great shape. We can talk about that maybe a bit later. And on the flash business, we, you know, through the downturn, we executed on a profitability basis better than anybody in the industry.
We have a very balanced portfolio across a great consumer franchise, client SSD, NAND, and we have an emerging enterprise SSD portfolio, which I think we'll talk more about, and we think that these businesses are just fundamentally mispriced. We think they're different businesses from an investment perspective. The HDD business is a lower capital intensity business, you know, more predictable cash flows. I think investors want something different out of that business. They want capital return policies, and the flash business is, you know, a more capital-intensive business, maybe a little bit more cyclical business, and we think, you know, when you look at what investors are looking to do in storage assets, we think these are better off in separate assets, especially now that they're executing extremely well.
I think that we started down that process. We announced it October thirteenth of last year. It's been a very big process inside the company to separate, and the company's got 50,000 plus employees. It's two big franchises. We're making very good progress down that path, and we expect to get to the point later this month, early October, where we've got all of our IT systems separated. We've got ourselves organized, and where we can start operating as two separate companies and make sure that we can, you know, process orders, make sure we collect revenue in the right way, get everything attributed to two separate businesses. It's called a soft spin in the parlance of people that do spins all the time, and we'll execute in that mode for the next quarter, through the December quarter.
And then when we get through that, we get conviction, we can close the books on two independent public companies. We'll do that most likely through the December quarter, and then we'll go through the separation process. So, it's been a very long process. We're very excited about it. We think we have two fantastic businesses. Gonna, you know, spin the flash business, give investors the opportunity to decide how they want to invest in those as separate assets. As I said, they're both executing extremely well. We're also making progress on continuing to build out the teams for each business. I think as we've talked about last quarter, I'll be going to lead the flash business. Wissam will be going to be the CFO of the HDD business.
He'll be teaming up with Irving Tan, who joined the company about three years ago and heads operations out of Singapore. He'll become the CEO of that business. I'm extremely excited about the two of them. I'm going to miss my partner here in Wissam, and we've had a great run together in this business, but I have an enormous amount of confidence in Irving and Wissam are gonna be great running that business, and within the last couple of weeks, we've hired a new CFO for the flash business. He's now joined the company, a gentleman by the name of Luis Bseiso, and he's got a very, very deep background in both the consumer business and the technology businesses, and I think one of the very interesting things about our flash business is we do have a very large consumer franchise.
About a third of the revenue is consumer, and, you know, very important brands like SanDisk and SanDisk Professional. And I think Luis joining us to team with me to run that business is another step in us getting to the point where we actually go execute this separation. So we're really excited about where we're headed.
... and, we think we're on a great path to unlock the value of these franchises.
The question I've been asked about Kioxia, that's your partner for the flash. They're going through some kind of a IPO on their own. Does that have any changes or any impacts to the JV?
The JV has been a great structure for many, many years, over two decades. I expect it to continue to be a great structure. In fact, in the last quarter, we've signed an agreement for the Kitakami Fab 2, which is our next production site, which will extend the JV through 2034. So, you know, a big part of the JV is our combined roadmap. You know, we work together on the BiCS roadmap, where, you know, we're, because of our combined share, we're able to invest an enormous amount of engineering resources in that. We think that puts us in the best position in the industry to have the best fundamental technology roadmap.
We look back over the last decade. We believe we've invested about a third less CapEx in the business to get this, get the output we need as industry average, and I think that's a reflection of our ability to drive an R&D roadmap that is very, very sound and very sophisticated. Quite frankly, we're able to invest as the market leader in having the best technology. We have the manufacturing behind that now in Kitakami through 2034, the Yokkaichi fab through 2029, 2030 timeframe. The JV has been through a lot of different corporate structures on each side of this, and I don't expect that to change at all as a part of them going public and our going through our spin.
It's just another iteration of the structure of each of our companies, but the JV is very sound, and it's a very, very big structure. Something that I've invested quite a bit in since I've been here, about making sure that relationship is very strong, and continues to be strong, so we expect that to continue to go forward. We are looking at, you know, as Western Digital, and we go to this spin structure, we do have this opportunity in this 2029, 2030 timeframe, where we have some of the capacity becoming available, that we could potentially move to a different fab or even a captive Western Digital fab. I think that's optionality we have, and anytime you have an optionality, you're always gonna evaluate what you're gonna do at that time.
So we're looking at that question, but, you know, having just signed an agreement through 2034, we're really, really, really happy with where the JV is at.
Great. Great. Question about end demand, you know, as you sit here in calendar third quarter, maybe if you can think about how you think the end demand across the various, you know, client, cloud, consumer business has changed, let's say, relative to where you guys were six-to-twelve months ago, and when you think about your end demand going forward, what are some puts and takes to your end demand outlook?
So twelve months ago was a very interesting pin to put in the calendar as far as where we were, because that's where we were kind of in the depths of the downturn, and we were just starting to see the business turn. The NAND business I'm talking about. And so we did see a very, very strong recovery in the business. Everybody has seen that for the last several quarters. And that was really, you know, one way to think about, led by the PC business. A lot of recovery in building back inventories and getting back a level of product into to support the normal flow of the business. That really led the recovery over the last several quarters.
I would say that business is taking a bit of a breather now, and maybe a little bit of a plateau before it makes its next move. And, you know, the market that didn't participate in the beginning was the data center market. We didn't see the data center market really moving at the beginning of the recovery. Now we're really seeing leadership transition to the data center market, and that market is very strong. So we've seen that transition in the business. Of course, around that, we had good consumer recovery in the first half of the year, and now we're seeing, I think, some general consumer headwinds and just consumer spending. So that business is taking a little bit of a breather as well.
But, you know, over the last, you know, two, three quarters, we're seeing that rotation and leadership of where, what's driving the market forward. When you look in the NAND business, you know, the big picture driver for us is always we look at CapEx spending across the whole business. Like, when is that? You know, how are people investing in the business? And, you know, we've been pretty clear that we're gonna look at the business differently. We wanna see, you know, sustained, better than through cycle margin target performance before we think about investing CapEx. And we're seeing the business get back into that through cycle and above level. So we're still in that, we're still in that phase.
When you look back the last couple of years, we haven't seen a tremendous amount of investment in NAND, so we expect continued supply-demand, good supply-demand dynamics in that business for increased profitability. But of course, it's a very dynamic business. You're gonna see different changes in different pockets of the business. So I said, leadership is transitioning from one part of the business to another. So that's kind of how we see the NAND business. HDD business, again, as you start to see that data center recovery, very strong in HDDs. We've talked about that the last couple of quarters. We just had record profitability in that business last quarter. We continue to see very strong demand for HDDs. As part of the downturn, we did kind of a major reset on the amount of capacity we have in that business.
I think for the first time in quite some time, we're seeing supply-demand balance in that business, which is leading to better dynamics around visibility and future demand on the business, pricing. So we expect that to continue. You know, we have good visibility over the next two-to-four quarters in that business. Do you want to add anything to that, Wissam?
I think you've covered it very well, David. The one thing I would add is, you know, you mentioned throughout the downturn, we took some cost action out, and so it's been, you know, we started seeing that margin expansion at lower volumes a few quarters ago. And as sort of the recovery continued, obviously, as David mentioned, we got to sort of record profitability with volume coming back, pricing being in good, also a tailwind for us. And we expect sort of we will continue to focus on margins and profitability going forward.
Fair enough. Digging into the flash business, maybe first, David, people are really nervous here about, you know, flash pricing. What's going on there? If you've just started sort of, you know, if you're just starting to get recovered here, why is flash pricing maybe across certain end markets not doing as well? So if you could just dial down on that one.
Yeah, the flash market is a very big market, first of all. Almost any interesting technology and device in the world is going to use flash, right? There's three huge markets in smartphones, PCs, and data centers, and then all kinds of things around that, around IoT devices.
Mm-hmm.
We all know how ubiquitous flash is in the world. We, I think it's what you know, I kind of outlined it a little bit earlier. Each market is going to behave a little bit differently, and they're going to recover at different times. They're going to, you know, as we saw, for example, the PC manufacturers, I think kind of OEMs, lead us out of very, very strong recovery in the flash market. I mean, a very strong recovery over the last couple of quarters.
You know, naturally, as they build back up inventory positions, they're going to get back to a level where they don't have to have quite as much demand after they build that inventory position and maybe take a little bit of a breather as the business hits a plateau before it takes the next move up. I think that's a natural thing that we're seeing. As you know, we didn't see the data center business move right at the beginning. We're seeing very strong data center demand now, and then you mix in with that macro headwinds or macro consumer, you know, spending around, you know, parts of our business where you have consumer and channel, more transactional business, and we're just going to see a lot of different dynamics around how each business moves and the pace it moves.
And it's not, you know, every business, every part of the business is going to be a little bit different each quarter. Now, when you add it all up, you know, we're still seeing like-for-like pricing increasing going forward. We're seeing volume increasing, and we're seeing our blended pricing go down a little bit this quarter because of mix. We're mixing more into mobile than we would in a previous quarter. So you're always going to see that kind of dynamic. But again, I'd back up and look at the big picture and look at capital spending and where supply-demand is at, and I think that's still in a very healthy place.
For as each of these markets continues to go through its own individual, you know, cycle, if you will, that the overall dynamics of the market, we think, are still in a very good spot.
Mm-hmm. Mm-hmm. Good. All right.
Talking about the end demand, you know, the question obviously, especially as it relates to PCs and smartphones that you talked about, is AI really an incremental driver for maybe additional NAND bits in each of those end markets?
We believe it will. I mean, we, we put out a webcast this past quarter on the AI data cycle, where we try and kind of demystify how AI really impacts the storage business. You know, there's many different phases of that, where it impacts both the HDD and the flash business. And the flash business is gonna gonna impact it in both kind of the training cycle of the AI data cycle, and we're seeing a lot of that with AI data centers being built and the strong demand for enterprise SSD. It's also going to implement the edge on devices, where you get to the inference phase and this starts to get deployed. So we're certainly believers that AI models are going to come to smartphones. The AI PC is going to be real.
We think smartphones are going to be impacted first as we come out of this. As the manufacturers prepare for that and put more capacity per device, we think the AI PC follows that. So we think that these are part of the good macro tailwinds of AI. You know, getting past just all of the training phase, getting into the inference phase, getting these models deployed on devices. And, you know, our NAND business is predominantly a device business, right? I mean, there's obviously a big data center component in it, but there's a huge device business out there for NAND, and we think as AI becomes a nice tailwind to that business.
Yeah. All right. And then the data center, you know, switching to eSSDs, can you—I know it's a growing part of your business, and you've talked about it being at least double digits for the fiscal 2025. So maybe you can just remind investors, you know, what are your current offerings there, how are you thinking about growing share in that business, and where are you seeing traction specifically as you're growing that business?
So I would say pre-downturn, we had just gotten to the point where we had qualified enterprise SSDs at a number of major hyperscalers just in time for the market to kind of really go into a major downturn and not a lot of buying happening in that market. So we're seeing a couple things now as data center recovers. One, we're seeing that traditional kind of hyperscale data center demand return. And then, so that's good. We already have products qualified. You know, those we're shipping those in more volume now.
You know, we also kept our, you know, shoulder to the wheel during the downturn and continued to build new products, and so we've just launched a new enterprise SSD focused around the compute part or that when you're going through model training, of really feeding that GPU complex, and so that requires a very, very fast PCIe Gen 5 enterprise-class SSD. We're just launching that product. It's in many, many, many, many qualifications. It's been very well-received. We believe it's the highest-performing enterprise SSD in the market for PCIe Gen 5. So that's really hitting this different customers building large language model training infrastructures. That product really hits right in that sweet spot of the high-performance enterprise SSD that's important for that. So we're seeing emerging traction there.
Of course, we're seeing demand for larger 30- and 60-terabyte enterprise SSDs. As customers build out their training infrastructures, they need those large enterprise SSDs as they prepare the data lake. You know, they pull all that information off of HDDs, they prepare it in a data lake, they do all of the things they need to do to get that data in a form where it can then be pulled into the training complex by these faster enterprise SSDs. And so that's we're building those higher capacity enterprise SSDs. We expect to have those in the market in the second half as we go through 2025. So we're seeing demand from the products we had qualified before the downturn. That demand is returning.
We're putting new products in that really hit the sweet spot of these AI training infrastructures, so we feel like the portfolio is well-positioned. As we said on our last call, which you referenced, we expect when we add it all up for fiscal year 2025, that enterprise SSDs will be about 10%, you know, double-digit % of our bit shipments for the year. So that's good progress in one year about where our enterprise SSD business is. You know, again, our portfolio, what we're trying to do in our portfolio is have a balanced strategy. We wanna be able to sell our products and our supply to as many places as we can, and then mix across all those different outlets to get the best return every quarter.
So as we talked about, you know, we have a great consumer franchise, and we can sell around the world. We have hundreds of thousands of brick-and-mortar places where people can walk in and buy our products and buy SanDisk products. That's a great franchise for us. We're very well-positioned in the client SSD business. We have a great position in mobile, IoT, auto, gaming. We have a great brand in gaming on WD Black. Also we're able to play that market with the gaming console providers and the retail side of it, and now we're bringing in the final major pillar to that portfolio strategy, which is enterprise SSD. And what we try to do every quarter is, you know, where do we see the demand across those markets?
Because as we talked about earlier, all those markets are not all gonna move in unison. Like, every market's not gonna be the same every quarter. They're just too big, they're too complex. Every market is gonna have a different demand, different profitability, and we wanna be able to mix across that to get the best through-cycle performance. That's what we're driving for, best through-cycle performance. And I think as we saw throughout the downturn and the recovery, that's what we're delivering, the best profitability on a through-cycle basis for the NAND portfolio.
Okay. A little bit on the node transitions, when you talk about the flash business. Where are you in the node transitions? Maybe if you can double-click on that, and how does that affect your market share positioning, or does it introduce any kind of risk to meeting demand?
I think from a nodal strategy, we have a lot of opportunity, right? As I said before, this is one of the, and I think that a lot of times in the JV, there's a lot of focus on the manufacturing side, and appropriately so. I mean, NAND manufacturing is important. But when I look at it, it's really that team that's defining the bits roadmap, right? And, you know, when you run a technology business, you think about how many engineers you can put on solving a particular problem, and the number of engineers you can put on solving a particular problem is almost perfectly correlated to your market share, right?
The fact that when you put us and Kioxia together, we're able to put, you know, that amount of resources on driving the BiCS roadmap forward, is why we have such a great technology base. I think if you look at this over a long period of time, we stay very focused on delivering the best technology at the least amount of CapEx required to bring that technology to market. Right now, we're in a position where we have a number of nodes kind of in production. We still have BiCS5, which has been the most capital efficient, most productive node in the history of the BiCS5 roadmap, is kind of the workhorse of the portfolio right now.
We have BiCS6, which we're producing, which is really, for us, is where we're gonna move parts of the portfolio to BiCS6. Usually, we would move the whole portfolio from one node to the next node, to the next node. I think as you heard, Rob Soderbery talked about the new era of NAND. We're thinking about this very differently now. We wanna move the portfolio where we get the best innovation out of each particular node. So BiCS6 is going to be where we're moving parts of the portfolio around QLC. We also have a new client SSD we've just launched in the last couple of quarters on which is a QLC-based client SSD. We believe that product has the best performance in the market on the client SSD side.
Really strong QLC performance out of that BiCS6 node. Then, of course, we still have BiCS8 in front of us, where we've talked about a lot, where we have this very significant architectural breakthrough on wafer bonding, or circuit bonding to the array, where we can produce the CMOS separately from the NAND, the NAND stack, and then we bond these two wafers together. That gives us an enormous amount of benefit. One, you can scale yields independently on the NAND stack and the CMOS wafer. The CMOS wafer, because you're not building the NAND stack on top of it, is, you know, maybe a little pure, and so the interface speeds you get out of that are just really, really superior.
We can now start to think about bonding different CMOS technology to different NAND technology to create different sets of products. So we can now have different innovation levers. Now, you can move those two levers at different rates and different paces. So, you know, that product, we're extremely excited about. It's, you know, it's starting to produce now. It's completely in front of us, so we think from a nodal strategy, we have a very strong lineup. We've got a lot of really great innovation that's in front of us, and as we wanna drive those nodes forward, is the demand is there in the market, and we see the profitability where we want it to be. We have all of that in front of us.
Of course, if you go to the R&D teams, they're off working on stuff that's well beyond BiCS8. We're not talking about that just yet, but again, that's that part where we have this very large R&D team between us and Kioxia, and we're looking at nodes well down the road from those. We think that this is a particular strength of the company. It's a particular... You know, in the NAND business, you have to start with good fundamental technology. You can't make it up by designing a controller on top of the if you don't have the fundamental technology in the right spot.
This idea to be able to innovate within the node, around performance, around density. This is really where the business is at, and we think we have a tremendous foundation under us, and the teams have really, really delivered as far as giving us a very strong NAND technology roadmap.
Okay. If I can switch a little bit maybe to the HDD side of the business now. You've talked about demand drivers here for nearline cloud customers. You know, just a little bit on visibility, how do you feel about visibility in that end market? And then people often ask this question: Does AI really an incremental driver for HDD side of the business, right? I know we talked about flash a little bit, but maybe help us understand why you think AI is incremental for HDDs as well.
Wissam, take this one.
Yes, sure. So on the visibility of the business in the nearline specifically, we do have visibility much better than sort of prior quarters. We can see within at least another maybe two to four quarters ahead of us. You know, coming out of the downturn, as we've resized our capacity, we've also implemented a build-to-order approach, which basically we work very closely with our customers to get earlier visibility of our demand, of their demand. And so, what we see is they're providing us longer term forecasts, which allow us to get better visibility to plan also our supply chain. So, yeah, as I said, we have visibility for two to four quarters.
With respect to AI, our current demand, what we see in our current demand is more of normal ordering patterns and inventory replenishment. We don't think, based at least today, we don't see necessarily demand driven by the AI side of things. Now, longer term, yeah, we do see AI as an additional sort of growth drivers that would enhance the CAGR or the longer term growth of the bits demand in the hard drive, on the hard drive side. But it's not yet visible in our numbers today.
Okay. All right. And then I guess, you know, there's always concerns about cloud customers, how close you are to their order patterns. Is there potentially any kind of double ordering going on there? I know we touched a little base, a little bit on that last night, but if you could just expand it over here for the broader audience.
Yeah, I mean, look, these are the, in any type of situations where we see demand sort of extending over a few quarters, there's always that kind of question that comes up, but we continue to triangulate with respect to trying to figure out where the end demand stands. You know, from where we stand today, we don't think there's double ordering happening. And obviously we will continue to monitor very closely, because that's also very important for us to be able to manage our supply chain. The way to think of building a hard drive, it takes almost...
Today, it's taking almost 50-52 weeks to build a hard drive, and so this is a very long lead time supply chain, and we need to be able to manage accordingly, and so this is why having that earlier visibility and demand, as I mentioned earlier, on the build-to-order is extremely important, because it's not only about our manufacturing process, it's also about the components that go into the hard drive, and we need to be able to supply those, and in tandem with our own manufacturing.
Fair enough.
Sorry, just one comment on that. I think that, you know, I'm really excited about where the HDD business is at. I think this business is just structurally a different business than it was several years ago, and I think, quite frankly, everybody involved in the HDD ecosystem is adjusting to that. HDDs were a technology. It's always been a very, very important technology in the data center. I mean, the foundation of storage in the data center is HDDs, and, you know, we don't see that changing anytime soon. They're both NAND and HDDs are great technologies, they're both gonna grow, but HDDs have always been plentiful, right?
HDD has always been available, so customers have been able to basically, at any time, be able to acquire HDDs because the industry is coming off of this literally 15-year transition from client to cloud. And in client, you just needed many, many more units. We don't need all those units anymore. And, you know, we've structurally changed our business to take that capacity out of the system. So, you know, customers, our partners can't just show up at any time and like, "Well, I need HDDs," and we have them available. Now, we all have to plan a little more, right? So we started with, you know, looking for some visibility.
I mean, just, I've only been in this business for four and a half years, and when I came into the business, like, the entire business transacted at the beginning of a quarter. Price was completely renegotiated every quarter, demand was given out every quarter, and so we've been slowly driving a change in that business, but coming out of the downturn, there was a structural change, and capacity was removed from the system, we think, to get supply-demand in balance. I think we're at that point now in the industry, where supply and demand are more, more balanced, and our customers, we're asking them to give us more visibility to what their demand is a year out, because we have to plan. We have to, you know, we have to put wafers through a fab, which is a very, very sophisticated process to build the heads.
That's why it takes quite a bit of time. We obviously have our own supply chain to manage for lots of different parts, and we've got to assemble the whole thing and test it for quite a bit of time before it actually goes into a data center. I think one thing that should be fairly obvious to everyone over the last... You know, watching the HDD industry over the last, let's say, 18 months, is the quality that's required to put an HDD in a world-class data center is a very, very high bar, right? It's not something that can be done very easily. There's a lot of work that goes on. The R&D is extraordinarily sophisticated, the amount of testing that's required.
You know, if somebody's gonna put that component in their data center and basically run it for five years and never expect to, you know, have any kind of problem with it, that's a very high bar, high bar. And so it takes quite a bit of, you know, sophistication to build a product like that. And I think we now have the industry is more balanced, our business is more balanced around supply and demand. So we're asking our customers to give us more visibility, and quite frankly, they've reacted quite well to that. Like, they understand. They understand it takes a long time to build this product. They understand that the product is very important to their infrastructure, and we're working with them to get that more visibility.
And at the same time, we've been able to, we think, you know, get a better share of the amount of innovation we're delivering. Remember, every generation of drive that we deliver to the market, we've just announced a 32-terabyte drive now with our UltraSMR and ePMR technology. That drive delivers better total cost of ownership to our customer. That allows us. You know, that total cost of ownership, there's a lot of value in that. And so we're able to, you know, drive pricing in a way that makes that product more profitable for us, and still be a great value proposition for our customers.
So I think that dynamic in the industry, it's been really, it's been really great to watch that play out and drive that over the last, you know, last two, three quarters coming out of this downturn.
Okay. Last five minutes, I'm gonna ask the audience if they have any questions. Please raise your hand so we can bring the mic over to you. Any questions from the audience? We have a question here, please.
Yeah.
Yeah.
Okay.
You're good.
Can you talk a little bit about the capital structure planning process for the two spin companies, or the split up of the company?
Yeah, maybe I'll touch on that. So, on the capital structures for the two companies, the way we're thinking about, and we're planning on, sort of architecting the capital structures is for each company, we look at the specificities of that company, meaning the markets it's in, the capital investment that's required, the growth profile, the volatility of the cash flow. All the various, and the volatility of profitability, of basically all the various things that you'd expect to look at by company, but also the competitive landscape, how sort of that asset would compare to the companies that are already out there trading in that space.
So all of these are inputs that we take into consideration, and our goal is to set up two companies that have competitive capital structures that can compete effectively and basically be successful in their own markets, but also, in many ways, be really good, attractive assets for for investors. So, as much as I'm not giving out numbers, I'm giving you sort of the framework of how to think about it. Now, as we get closer to the spin, obviously, there'll be our SEC filing becomes public. That will have a pro forma of the cap structures in it. And also that would be made available ahead of an investors day as well as roadshows.
So there'll be opportunities for investors to look at the material, be ready for investors day to ask sort of the good questions around the how this whole thing was made up. So hopefully that helps.
Any other questions from the audience? One more.
Can you talk about the HAMR technology and then the timeline in your roadmap?
Yeah. We've, I think maybe for the first time in a very long time, different players in the industry have different technology roadmaps, and, we very clearly made a decision quite some time ago, quite frankly. And I think I've said this before, our chief architect showed me some material where it was all the way back in 2013, I believe, that, you know, the team looked at this and decided we wanted to develop the ePMR and the UltraSMR technology, before we got to HAMR, and that's what we've been doing. And we've been working on HAMR for a very, very long time. HAMR is a big part of our roadmap.
We, you know, because of the technologies we've developed around ePMR and UltraSMR, for us, HAMR is not accretive to the portfolio until we get to forty terabytes per drive. You know, to build a HAMR drive, it's a different bill of materials, different components. It's just more expensive. So you have to have, you know, you have to have a big enough capacity point to amortize that cost for us to get accretive. And so, you know, we've got, you know, we built a set of technologies around ePMR, OptiNAND, UltraSMR, that allow us to drive our roadmap up to forty terabytes on that technology, and that's exactly what we're doing. And that technology is proven. You know, it's easily qualified by our customers.
That thing I said earlier about how difficult it is to build a hard drive that goes into a, you know, production data center in the world is a very, very difficult task. It's a very high bar. It's not like building a hard drive ten years ago that goes into a desktop system. It's a very different set of requirements in the operational requirements, the environmental requirements, the reliability requirements. It's a very different device than it was ten years ago. So, we built a set of technologies that'll our customers understand extremely well. It's been qualified. You know, products are. You know, we ship millions of these products a quarter, and, you know, we've just announced the 32-terabyte new drive, UltraSMR drive. The industry has adopted SMR, quite frankly.
I think if you go back two years ago, there was a big question around that. People saw SMR as something that was mainly a consumer technology, and it wouldn't be adopted by the hyperscalers. That's, you know, I think we've driven the market to adopt SMR. We have three hyperscalers now, or two hyperscalers now on US hyperscalers on some variant of SMR, and we have the third one adopting it, you know, aggressively as we go into the second half of the year. So it's very clear that SMR is the current and next generation of technology in the world-class data center, and we're extremely well positioned for that. Our UltraSMR technology gives us, you know, 20% additional capacity over a CMR drive. Everybody else in the market is a 10% amount of capacity increase.
That 20% comes from the fact that we have OptiNAND in there, and we'll be able to do, you know, do some interesting design work and software that gives us another, that incremental 10%. And that's really important, right? It's really important for the customer because they get a drive that's denser. It's a better TCO equation. It's really important for us, quite frankly, because we're able to get that extra 10% of capacity without adding any additional cost into the system, so it's a more profitable drive. And I think you're seeing that in the industry, where our profitability is now separated and, you know, is increasing and is putting us in a very good position. Look, we're big believers in HAMRs.
HAMRs, they will come, but it's a very high bar to get into those data centers, and we have technology that's gonna drive us to forty terabytes, and at that point, we'll make the transition to HAMR, and we expect that to be a seamless transition.
Cool. Great. Thank you. It's-
Thank you, everyone. Appreciate your time.
Thank you, everyone.
Thank you for having me. Thanks, everyone.