Let's get started. Good afternoon, everybody. Welcome to the Goldman Sachs Communicopia and Technology Conference. My name is Jim Schneider. I'm the Semiconductor Analyst here at Goldman Sachs. It's my pleasure to welcome Sandisk today, CEO David Goeckeler, CFO Luis Visoso. Welcome, guys. Thank you for being here.
Thanks for having us. It's great to be here.
David, maybe start with you. It's been roughly...
Yeah, it's always hard at first.
We will be making forward-looking statements in today's discussions based on management's current assumptions and expectations, including with respect to our product portfolio, business plans and performance, mass market trends and dynamics, and future financial results. These forward-looking statements are subject to risk and uncertainty. Please refer to our annual report on Form 10-K and our other filings with the SEC for more information on the risk and uncertainties that could cause actual results to differ materially from expectations. We will also be making reference to non-GAAP financials, and a reconciliation of our GAAP to non-GAAP results can be found on our website. Thank you, Jim.
Very good. David, maybe start with you on some high-level strategy questions if we can. It's been about six months since you executed the split between yourselves and Western Digital.
How have you progressed compared to your initial set of expectations you laid out at your investor day?
First of all, it's been a quick six months. It's been quick in some sense, and it seems like a long time ago now. I think the expectations we set out at our investor day back in February have been pretty much true to what we've been delivering. We're just really, really optimistic about the market we're in, the potential for the market we're in. We can talk some about supply-demand dynamics, which I'm sure we'll get into. We think we're in a long arc here of the industry getting to a point where supply-demand is balanced and tilted towards an undersupplied market. We've believed that for quite some time. We think the portfolio is in fantastic condition. We talked about our new enterprise SSD programs at our analyst day. That program continues to play out extremely well.
We now have that product in customers' hands to start initial qualifications, so that's great. Our client portfolio has been strong for a very, very long time and continues to be that way. Of course, our consumer brand has always been a strength of the company and we continue to invest well there. Inside the company, the team is just incredibly excited. I think to be back as Sandisk, as an independent company, has been a big jolt of energy for everybody, and we're just really, really happy with where the company setup is. We kind of came through the separation just flawlessly and just continue executing along. I'll let Luis give his point of view as well.
Yeah, I think it's been very interesting. We have two quarters that are behind us, and we deliver everything we promised we would do in February, so we feel very good about it. I think the one metric that makes me very happy is our cash flow generation. I believe cash at the end of the day is what matters the most for our shareholders, and we're pretty much in line with what we expected. We've been reducing our debt. Our net debt is pretty much in good shape, and we should be net cash flow positive soon.
Great. Six months on, anything changed about the way you intend to run the business going forward that you want to course-correct on?
I think, again, we had a thesis on the market when we came out and the way we set up the company, and I think that that's essentially played out. We're just very optimistic about this business and what we can, you know, where we can drive it to and the value we can create. There hasn't been any need for any course corrections. What we're doing, as I said, we think the portfolio is in great shape. We can talk about our fundamental NAND technology. We're now going to ramp into BiCS 8 over this year. We're going to go from a single-digit percent of our portfolio on that technology, and we'll be 40%, 50% by the end of this fiscal year. Our customers are telling us it's the gold standard for nodes in the market right now.
We just feel really good about where we're at and what we're driving towards and our ability to generate value with this franchise.
Great. David, you've clearly articulated the view that supply-demand dynamics in the NAND industry are improving. You see an undersupply situation lasting into next year. What catalyzed this change, do you think, after quite a long period of weak industry dynamics? Do you believe the change in competitor behavior is sustainable from both a supply and pricing perspective?
I think that's a very complicated question, but let's decompose it a little bit. First of all, on just supply-demand, obviously, we talk to a lot of customers. We have a lot of visibility. We have a big consumer franchise. 20% of the clients in the world use our technology. We're a player in the enterprise market as well. We do a very, very detailed bottoms-up planning of, you know, starts with the three big markets: PC, smartphones, and data center. If you look at those markets going back, you know, even a year ago, we were becoming bullish on this setup because we saw everything going in the right direction. Smartphone units up slightly, content per device going up. We saw that in the Android ecosystem first as they prepared for AI on the devices.
Yesterday, we saw it from another big player in the business of kind of upping the content per device. We see PC units growing and content per device growing. The other big market is data center. We've seen nothing but improving dynamics in the data center. We're coming off of a year of 24% of 130% exabyte growth in that market. On top of that, 25%, we'll see double-digit growth. That number keeps going up. We had an earnings cycle a couple of weeks ago. That number went up again. We see that number even stronger, that growth increasing even more in 2026. We've been looking at these markets for quite some time. We've seen the way this setup is coming out. We've just been very, very bullish on the demand side for NAND. Then we look at the supply side.
We all came through a very, very stressful downturn where I think the traditional way that you manage this business of saying it's an elasticity-driven market. When you start to see unfavorable financials, you release a new node that drives your costs down, and the market is elastic enough to absorb all of that supply. That strategy ended in the last downturn, and it ended rather spectacularly. Right. The way you manage the business now is just fundamentally different. We've got to spend more time being focused on making sure supply is matching demand. The very good thing in NAND is R&D is alive and well. We can produce more NAND. We have a roadmap for more density every node we produce. That's a good thing. We got to make sure we don't turn that knob too far, too fast.
If we keep those two things in balance, we will see very good financial dynamics in this industry and kind of turn around what's been happening over the last at least five and a half years that I've been a participant in the industry. That's naturally happening. It's because I think if you run one of these franchises, you look at these financial dynamics and you come to the same conclusion. If you look at the amount of CapEx that's being spent in the 3D era, we've kind of dramatically reduced that over the last four or five years. Nodes are more expensive. They're more productive. We need to pull back on the level of capital intensity to keep supply and demand balanced. We've liked the setup for a long time. It's a big market.
It takes a while for all of that to work its way through inventory at suppliers, inventory at customers, different markets going up and down. We don't see a good market going into 2026. We see an undersupplied market all the way through 2026.
Do you have the same level of confidence in the sustainability of your competitors' supply?
I can only run our business, right? We run our business, and we're very focused on profit generation, cash flow generation. I think we're doing very, very well. We're going to drive the business to net debt neutral to positive here very quickly, going to get to the point where we need to be. I think if you run one of these franchises, you're looking at these dynamics and realizing we just can't keep putting new nodes out all the time and spending the amount of CapEx. 3D era nodes are 2-3x more expensive. The good news is we've got a lot of R&D productivity. We need to make sure we invest in that at a sustained rate. I'm a believer that this is going to be, it is and will be a better industry, and this franchise is enormously valuable.
Great. There was a press report recently that you are raising pricing on consumer and shell products by about 10%. Can you confirm that and maybe contextualize that in terms of your broader pricing expectations that you've already laid out?
Yeah. We announced a price increase for a part of the market. As you mentioned, these are the customers we don't talk to directly. Our expectation, as David said, is the market is tight, and we'll continue to see opportunities to increase prices across the board. If you think about our client business, we negotiate every quarter with each of our customers on volume and price. We're going to go through those processes. We're going to do the same on the data center side, on the cloud side. It's going to be based on the market dynamics on where we end up.
You should not extrapolate that 10% increase across the board, but that's an indication of where the market is going.
Yeah. How should we think about Sandisk's overall bit supply growth this year and next?
Look, we have a very clear stated goal to grow with the market. I mean, it's a market where things change kind of slowly on a share basis and growth basis. We're in a mid-teens CAGR demand growth in this market, and that's where we expect to grow in line with that.
Got it. Have those dynamics changed at all relative to what you thought before? Do you believe that there's any consolidation needed in this industry beyond what's, or do you think we're kind of in good shape and we can kind of sustain it with the current number of players?
I definitely think that the market is, I think it's a great market to be in. I think that the dynamics of the market are changing before our eyes for the reasons I talked about earlier. You know, NAND is a business that's changed a lot. It was a business where we were really trying to drive growth in the market. We're $65 billion going to $100 billion. I think we can check that box. I think the way you manage the business is different, as we talked about. I think that is sustainable. That's the way we're going to run our business. It's kind of a separate conversation of consolidation. We're very happy with our business. Obviously, in a market where you have high fixed costs, any kind of consolidation is going to drive better cost dynamics. That's always welcome.
It's a complicated question when you get into the practice of it.
Yeah, I understand. Maybe a couple of product questions if we could.
Sure.
Think back to the Flash Memory Summit last month. You made an interesting announcement about the notion of high bandwidth flash called HBF. I believe you're targeting sampling sometime back half of next year. Maybe provide a little context for what that is, what's its significance, is it evolutionary, revolutionary, and maybe sort of how critical this is to future AI training and development.
This is one of the reasons I feel there's so much excitement inside the company, because not only are we really happy with where we're at in the core business and we've got, like, BiCS 8 is fantastic and we're developing new nodes and the portfolio is in the best shape it's ever been, we're also working on this kind of incredible technology of high bandwidth flash. Take a little bit through the journey of where this came from. I mean, obviously, anybody in the world today that has a technology franchise is thinking about how they play in the AI architecture. You know, one thing about NAND is we have a lot of density, right? We can store a lot.
In a world where models are getting bigger and bigger, and you know, we're trying to think about how do we apply our technology into that inference phase, we needed to focus on kind of how do we increase the bandwidth of NAND. We're not trying to build a DRAM substitute. We're not trying to build HBM. What we're trying to do is say in the inference phase of AI, which is going to have to go to devices, right? Again, 25% of every device in the world has our technology in it. How do we take NAND and apply it to this AI architecture? You know, if you're a NAND designer, you've been focused for the last 25 years on how do I deliver better density in NAND. You've done a fantastic job, like incredible job, like almost too good of a job, right?
We can deliver a lot of capacity, a lot more bits per wafer. Some of our engineers were looking at this problem and said, you know, what if we focused all of this intellectual energy on how do we make NAND faster? Could we solve the use case of inference for in the AI architecture? We knew we had one big part of the problem solved. We had density, right? DRAM has a different issue. It doesn't really scale anymore. They're doing a fantastic job. Nothing great for DRAM. If we could bring NAND to this equation, we could really be in a good position because we can bring customers an inference solution, much smaller footprint, much more density, much better power efficiency, all of these kinds of characteristics. We got the team working on that for the last couple of years, how to think about how to re-optimize NAND.
We came up with some really clever ideas. Then we went out and started to talk to some very big customers and say, how are you going to build inference architecture in the future? We started working together. That's where this whole idea of High Bandwidth Flash came from. We're very optimistic about it. We felt at our investor day it was time to talk about it publicly. After we talked about it publicly, we got approached by one of our peers that says we want to work together to standardize this for the industry. We thought that was a great idea. We announced that at FMS. We also put a timeline behind this where we said, look, we're going to have the NAND available in late 2026, and we'll have the system available in early 2027. The system is building the controller, building the ASIC for all that.
We're doing work on both of those. We're talking to customers about it and how they would deploy it in their infrastructure or on their device for AI inferencing. We're super optimistic about cracking that over the next couple of years.
Good. Is that just meant to compete with anything, or is it just meant to compete with on-device storage and just provide better performance?
I would say it's meant to enable AI inference everywhere. It's on your smartphone, on your PC, and even in the cloud, right? The cloud is going to have an infrastructure for inferencing, an infrastructure for model training. That infrastructure for inferencing is wide open territory for us to come and provide a solution that's much more scalable, much more power efficient, smaller footprint to drive inferencing. That's what we're working on. We're working with customers to understand what exactly are the inferencing use cases going to look like two years from now, three years from now, so we make sure we build exactly the right product.
Okay. There's been a lot of focus on the enterprise SSD market. You've talked about it quite a bit. Maybe remind us how qualifications are going for your products in that space. What are you hoping to achieve by the end of this year and into next?
This is an area where we've been building out our portfolio. We have a set of products that we qualified over the last two, three years. That set of products is deployed out there. When we talk about, you know, this past fiscal year, about 13% of our bits that we shipped were in enterprise SSD. That was that set of products. Now we have a whole new set of products coming out. There are really kind of two major products. We call one kind of a compute-centric enterprise SSD that's been in the market for about a year now, qualified at one hyperscaler, undergoing qualifications next. It's the product that was certified by NVIDIA. It was part of their reference architecture, was the first enterprise SSD that went through that process. That product's going well, and we're going to be ramping that over this year.
We have the product we talked about at our investor day, what we call our Stargate platform, which, by the way, we picked the names two, three years ago. It's now been reused, which is great. That's the high density for AI data lake. Start with 120 TB, 256 TB, going to 512 TB and eventually to a petabyte. That product has been in development for the past three or four years. We're just at the phase now where we're putting it in customers' hands for the first qualifications. You asked me the first question, has anything changed since our investor day? We're right on track with that program. We feel really good about where it's at.
We're going to go from a position of, we're going to be in the position to lead that transition to those next capacity points, be right there as the industry transitions to that and enable that. We feel really good about where we're going. This is going to be a story that plays out over years, right? This isn't a story that's going to change. It'll get incrementally better quarter by quarter, but it's really going to be a couple of years as we get these products qualified, expanded across customers, and deployed at scale.
It's going well, and we want to be in a position where, you know, we manage our portfolio in a way, great consumer franchise, great client franchise, great enterprise SSD franchise, gaming, and then whatever each quarter brings, how do we mix across all of those parts of the portfolio to get the best financial outcome?
Great. At the end of the day, I'm still a chip guy at heart, so I want to ask you about BiCS 8. You raised it before, the step forward in terms of, I guess, performance, but most importantly, cost reduction. Maybe remind us of your expectation for the like-for-like cost reduction BiCS 8 brings once fully ramped. Maybe we'll follow on.
This is one of the things we said at our investor day we're going to stop talking about. I appreciate the question, but we're not throwing out like-for-like numbers out there. I said at our investor day, it's kind of an interesting industry where you talk about your costs all the time, and then you go negotiate with customers on what the price is, and they know exactly what your costs are. You can assume that the like-for-like costs are lower, right? We're going to keep that for ourselves. It's, you know, we're going to go through a major ramp here for the first time in a while. We were on BiCS 5 for quite some time. It was a great node. The peak yields were fantastic, the best in the whole BiCS family.
We went to BiCS 6 for QLC on part of the portfolio, and now we're going to drive everything to BiCS 8 over the next couple of years. We have a lot of confidence in that node. As that ramps this year, that will provide some cost tailwinds to us. If you look at our business over the last year, we've been preparing for that. We had some fab start-up costs. We had tools we were buying to get ready for that transition. We threw in a little bit of underutilization costs along the way. We've had cost headwinds for the last three, four quarters in the business. Going forward, those headwinds are going to turn into tailwinds. That's why we like the setup of the business. Portfolio in good shape, supply-demand dynamics in good shape. We're in a market where we think pricing continues to inflate.
At the same time, our business is turning from cost headwinds to cost tailwinds.
Great. I want to ask about, oh, sorry, relative to BiCS 8, maybe just give us any kind of framing as for when you sort of achieve crossover from a production perspective.
What we've said is we're currently mid to high- single digits percentage of the portfolio. We expect to end this fiscal year about the 40% to 50%. That's when you'll see that crossover.
Got it. Maybe returning for a moment to the SSD discussion we just had, I think, you know, in the past, you sort of had noted a target or an aspirational level about 16% as the SSD market share. You know, where do you think you are now, and when do you think that target becomes realistic?
I would say right now, from a share percentage, we're mid to high- single digits in that territory, less than our share of bits. The way I think about it is we want to get to that share of our bits as a next milestone. We want to have the optionality to mix when it's the right thing to do, right? I mean, enterprise SSD is a great market to be in. It's great to run a franchise, quite frankly, where you have a great market to grow into. That's actually a good dynamic, a good tailwind from a mix perspective. That's what we want to get to. What each quarter brings, we'll figure out what the right mix is from a share perspective. I mean, remember we have this awesome consumer franchise, and on a through cycle basis, that is the best franchise out there.
We want to make sure that we get the right balance across the portfolio. Clearly, we're putting ourselves in a position where we'll have more optionality and drive that enterprise SSD share higher. The way you do that is you build great products. It's just as simple as that. That's what we're doing. The generation of products we're just putting in customers' hands right now, we have an enormous amount of confidence in.
Great. I don't want to make Luis feel left out. Let's talk about operational and financial trends for a second. I think, you know, relative to your spending targets, you know, given the industry's more prudent supply-side behavior, how should we be thinking about CapEx in terms of spending growth or net on a go-forward basis? Is the primary driver of that still the sort of BiCS 8 investment?
Yeah, so our model is to spend CapEx only in the mid-teens on a gross basis. This year is going to be a little bit above that. Why? Because we're transitioning to BiCS 8, which David has explained. As we continue to bring new tools and get them ongoing, we need to spend the CapEx to do that. Still, very importantly, our free cash flow is going to be positive this quarter. We said the free cash flow for the full year will be positive. Despite all these investments or incremental investments, we'll be free cash flow positive for the year so we can afford it. That's overall where we are on CapEx. On a net basis, it varies by quarter, right? You have depreciation, you have subsidies, you have leasing. It varies a little bit by quarter.
We feel very good about our CapEx plans enabling our innovation while still generating free cash flow for our shareholders.
Great. I think it's fair to say that your JV with KIOXIA is one of the few successful tech JVs over the past couple of decades. If you were to point out two or three key highlights of that joint venture, what makes it what it is?
You're right. It's a great relationship. I joined it five and a half years ago. It's been going on for over 20 years at that time, and it's just an awesome relationship. We're in a business, we're in a technology business. You know, you think about how you can invest, especially R&D. We talked about earlier, why is BiCS 8 a great node? Because we are able to invest with KIOXIA as much as anybody else in the industry. Together, we're largest or close to tied for largest of market share in the industry. You can essentially afford to invest commensurate with what your market share is in a business like this. I think the JV, a lot of people think about the manufacturing side of it, which had its own benefits, don't get me wrong. The R&D side of it is just of paramount importance.
The teams work together hand in glove. You would think it was one team. Developing NAND is not easy. I think sometimes people conflate a commodity-priced product with a commodity. They're very different things. NAND is far, far, far, far from a commodity. It is extraordinarily difficult R&D. The fact that we can work together with our partner, and it's our combined investment that's going into building that roadmap, that's how we end up with, you know, 10 years, a 10-year track record of spending a third less capital than the industry average. Our R&D team is very focused on how do we build our technology in the most capital-efficient way. It's how you end up with a node like BiCS 8 where you have wafer bonding, the first company to deliver wafer bonding at scale.
You look at the performance we're going to get out of QLC on that product is just fantastic. That side of the JV, I think, is sometimes a little bit underappreciated in a big part of it. Of course, you got the manufacturing, right? We manufacture together, and we have two incredible campuses in Japan at Yokkaichi and Kitakami that give us scale there. That's just a little bit about the JV. It's just a wonderful relationship.
Great. Back to financials, two last ones maybe to end on. Your gross margin guidance long-term model is 35% through cycle. Can we maybe unpack what through cycle means over time? Is that a true average you can expect to achieve? Is it a floor? How should we be thinking about that?
Yeah, it's clearly not a floor. It's clearly not a ceiling, right? It's an average that we expect in somewhere in the three-year average, right? Since we've been below that, we need to be above that, right? Over the next few quarters, what we should expect is, as we mentioned, we should expect some gross margin expansion. What we're seeing is the benefit of ASP increases. We see some of the benefits on cost, and we'll keep on driving that. We will see better gross margins over the next several quarters.
Yeah. Your long-term model also implies 15% OpEx intensity. How do we think about sort of your allocation of R&D relative to SG&A? Is there a desire to keep, you know, to grow R&D materially and try to kind of keep SG&A as flat as possible?
Yeah, the vast majority of our OpEx is R&D, and that's where it should be. I mean, David talks about this as an innovation company. Our innovation is the lifeblood of this company. I mean, it's just the most important thing, right? We have important investments to make, particularly HBF. If you think about ESSDs, we need to enable all of that. You know, we incur some costs as we launch our products, right? There are samples to our customers that we need to incur. We should get some of those costs, you know, sometimes during the quarter. We'll keep on investing in innovation. At the same time, we're driving efficiencies in SG&A. Wherever we can, we're moving people to lower cost locations where we can.
We're trying to do both things at the same time, fuel the business through innovation and R&D and driving efficiencies everywhere else where we can.
How do you think about the debt load and trajectory of debt paydown? To the extent you're generating very good free cash flow, if everything plays out as you say and gross margins and so on are running above normalized, would you accelerate that debt paydown?
Yeah, I mean, we've been paying our debt earlier than what we had to, right? We paid down $200 million, and we'll keep on doing some of that. At the end of the day, we want to be net cash positive. Once we get there, we're going to do three things consistent with what you said in February. We're going to continue to invest in the business. This is a growing business that needs some cash, and it's generating cash. We'll continue to invest in the business and continue to generate cash to reduce our gross debt. We will also return some cash to investors, right? I mean, that's what cash generation is for. We still need a few more quarters to get there, but that hasn't changed.
Yeah, just to confirm, sort of tactically, relative to the back half of this year versus what you had talked about at your investor day, it sounds like everything is on track for kind of tracking to your expectations versus prior?
I think at our investor day, we gave a very clear view of what we see as the market through the end of the year. I think it's playing out that way. I think we're happy with where it's at. As I said, the portfolio is in great shape. We're very bullish on the business. We're very happy to be here. I think, like as you started, separation went flawlessly. I think we're off and running. I think we have an awesome opportunity we're going to take advantage of.
Great. Maybe very last question, which is you've done a lot of investor meetings over the past day, and I'm sure over the past month or so, you know, when investors who are new to the story or ones who have been looking at this story talk to you, what do you think is the one thing that is underappreciated about the Sandisk story today?
I just think in this market, there's, you know, I experienced this for five years in the HDD market. There's just a very strong desire to want to map the past to the future. It's a natural thing to do. All these things are going to happen and the way things work. I think what you're seeing is a market that is fundamentally changing. As I touched on a little bit earlier, the fundamental way you think about this market, the fundamental way you manage this market, I think is changing. I think it's changing before our eyes. It takes a while. I think when you're in the middle of it, it's hard to see it. It's definitely happening. I think that's going to change the economics dramatically on the other side of this. We're seeing that play out in the HDD business, quite frankly.
I think we're going to see it play out in this business. This is a much bigger business. Bigger levers, bigger prizes. It makes me very, very excited to be a part of it.
I think it's a great place to end. David, Luis, thanks for being with us.
Thank you very much. We appreciate it.
Thank you.
Thanks, man.
Okay.
Thank you.
Appreciate it.