Seagate Technology Holdings plcSTXNASDAQ
LOADING
|||
Operator: Welcome to the Seagate Technology Fiscal First Quarter 2026 Conference Call. [Operator Instructions] Please note, this event is being recorded. I would now like to turn the conference over to Shanye Hudson, Senior Vice President, Investor Relations. Please go ahead.
Shanye Hudson: Thank you. Hello, everyone, and welcome to today's call. Joining me are Dave Mosley, Seagate's Chair and Chief Executive Officer; and Gianluca Romano, our Chief Financial Officer. We've posted our earnings press release and detailed supplemental information for our September quarter results on the Investors section of our website. During today's call, we will refer to GAAP and non-GAAP measures. Non-GAAP figures are reconciled to GAAP figures in the earnings press release posted on our website and included in our Form 8-K. We've not reconciled certain non-GAAP outlook measures because material items that may impact these measures are out of our control and/or cannot be reasonably predicted. Therefore, a reconciliation to the corresponding GAAP measures is not available without unreasonable efforts. Before we begin, I'd like to remind you that today's call contains forward-looking statements that reflect management's current views and assumptions based on information available to us as of today and should not be relied upon as of any subsequent date. Actual results may differ materially from those contained in or implied by these forward-looking statements and are subject to risks and uncertainties associated with our business. To learn more about the risks, uncertainties and other factors that may affect our future business results, please refer to the press release issued today and our SEC filings, including our most recent annual report on Form 10-K and quarterly report on Form 10-Q as well as the supplemental information, all of which may be found on the Investors section of our website. [Operator Instructions] With that, I'll hand the call over to Dave.
William Mosley: Thanks, Shanye, and hello, everyone. Seagate delivered a very strong start to fiscal 2026. Revenue grew 21% year-over-year. Non-GAAP gross margin set a new company record at 40.1% and non-GAAP operating margin climbed to 29%, a level last seen in fiscal 2012. Non-GAAP EPS exceeded the high end of our guidance range, underscoring our focus on expanding profitability. Today, we announced an increase to our quarterly dividend of approximately 3%, reflecting confidence in our execution and ongoing sustainability of our cash flow generation capabilities as we leverage our leading HAMR technology and a strengthening demand environment for high-capacity hard drives. Demand strength was led by global cloud service providers, and we also saw meaningful sequential revenue growth from enterprise customers in the September quarter. The data center end market, which is comprised of nearline sales into cloud, enterprise and VIA customers represented 80% of overall revenue. Amid this improving demand backdrop, our high capacity nearline production is largely committed under build-to-order contracts through calendar 2026. Additionally, longer-term agreements that we have with our global data center customers provide clear visibility through calendar 2027, reinforcing our view that these favorable demand conditions will persist. We remain focused on executing our HAMR-based product road map to support our customers' growing exabyte needs and continue working with them to transition to higher capacity drives. There is no question that AI is reshaping hard drive demand by elevating the economic value of data and data storage. This is evident by the growing demand for our high capacity nearline drives as customers continue to ramp investments in AI applications. AI inferencing is set to inflect and scale rapidly, further increasing data's value. Inferencing consumes and generates large volumes of data, which is then stored, monitored, validated and reintegrated into an infinite training loop. We are already seeing the positive impact of this trend as global CSPs deploy large-scale inferencing applications that rely on multimodal inputs such as text, audio and video. Using monthly token consumption as a proxy for inferencing adoption, one major hyperscaler reported a 50-fold increase in the span of a year. This explosive growth is driving a sharp increase in unstructured data generation that creates demand for hard drive storage. Video content is a major contributor of unstructured data and is driving considerable demand for hard drives today from social media platforms to content delivery networks and online marketing. AI-generated videos promise to further fuel demand growth. There are already numerous text-to-video tools that democratize creativity by letting anyone generate professional quality videos from text, images or sketches. We see this trend already taking hold. For example, Google reports over 275 million videos were generated on its Veo platform within the first 5 months. With a 1-minute AI video being up to 20,000x larger than a 1,000 word text file, the data storage implications are clear. The rapid adoption and growing capability of these tools are already having a positive impact on the demand for storage. Beyond the application space, we have discussed new storage use cases from emerging trends around hybrid cloud environments that enhance data security and compliance with data sovereignty regulations. Recently Seagate partnered with a global CSP to develop a sovereign cloud solution for managing massive volumes of sensitive telemetry and sensor data collected from a fleet of autonomous vehicles. These type of data sets are subject to strict requirements and stipulate such data must be processed, stored and managed locally. As in any other data center, hard drives provide the ideal solution by meeting customer requirements for throughput, durability and cost-efficient long-term data retention. As data generation explodes and new use cases emerge, Seagate is answering the call with a clear long-term road map to capture demand. Momentum continues to build for our HAMR-based Mozaic platforms, and we achieved several important milestones in the quarter consistent with what we discussed during our analyst event. We now have 5 global CSPs qualified on Mozaic 3+ terabyte per disk products, which can deliver capacities up to 36 terabytes per drive. We remain on track to qualify the remaining 3 global CSPs within the first half of calendar 2026. Additionally, we shipped over 1 million Mozaic drives in the September quarter. These products are performing well in live production environments, and we are on pace to achieve 50% exabyte crossover on nearline HAMR drives in the second half of calendar 2026, and we started qualification with a second major CSP on the Mozaic 4+ terabyte per disk platform, with initial volume ramp starting in the first half of next calendar year. This platform will offer capacities of up to 44 terabytes. Advancing aerial density is a key competitive advantage, not just for Seagate, but for the hard drive industry overall. We are leveraging our manufacturing expertise and advancements in technologies, including silicon photonics to pave the path to 10 terabytes per disk. Our aerial density road map delivers a superior and sustainable TCO advantage for hard drives compared to alternative technologies well into the future. Customers clearly see the value of transitioning to higher capacity HAMR products as the most efficient way to support their rapidly expanding data storage needs in an AI-driven world. Wrapping up, the Seagate team continues to execute at an exceptional level. We are delivering on our target financial framework supported by a structurally improved business and a strong sustainable demand environment. We are advancing our HAMR led technology road map, which creates significant value for our customers and positions Seagate for long-term success. With the strength of our technology road map and the transformative impact of AI, we believe the best years are still ahead of us. I am proud of how our teams are rising to meet the opportunities ahead as we remain focused on delivering profitable revenue growth and expanding cash flow generation in fiscal '26 and beyond. I'd like to thank our employees, supply partners and customers for their many contributions to our performance and to Seagate's ongoing success. Let me now turn the call over to Gianluca.
Gianluca Romano: Thank you, Dave. Our September quarter performance demonstrates strong operational execution and underscores the enhanced structural economics of our business model. We delivered revenue of $2.63 billion, up 8% sequentially and up 21% year-over-year. We achieved a record non-GAAP gross margin of 40.1%, up 220 basis points sequentially. And we expanded non-GAAP operating margin by 280 basis points to 29% sequentially. Our result in non-GAAP EPS was $2.61, exceeding the high end of our guided range. We have continued to execute our technology road map to support ongoing demand momentum for our higher capacity products. In September quarter, we shipped 182 exabytes, up 32% year-over-year, with the vast majority of that volume delivered to global data center customers. As we shared last quarter, we will be discussing the business across 2 key markets: data center, which is comprised of nearline products and system that are sold into cloud, enterprise and VIA customers. And edge IoT, which includes consumer and client-centric markets, along with network attached storage. In the September quarter, data center revenue represented 80% of our total revenue at $2.1 billion, up 13% sequentially and 34% year-on-year. Demand from global cloud customers continue to grow, and we also saw a notable improvement in the enterprise OEM markets. We project these positive trends to continue with cloud growth expected to outpace enterprise demand. Whether data is stored in public cloud, private cloud or on-premises, the shift from AI model training to inferencing is driving the need for large capacity or drive storage. This includes everything from saving checkpoints to maintain model accuracy and integrity to storing the vast data sets required for effective inference results. In the September quarter, we shipped 159 exabytes into data center customers, up from 137 exabytes in the prior period. Cloud exabyte demand increased for the ninth consecutive quarter, resulting in close to 80% of nearline volume on drive capacities at or above 24 terabytes as customers continue to mix up to higher capacity drives. Over the past year, average nearline drive capacity has increased by 26%, which is a primary contributor to our exabyte volume growth. Amid tight supply condition, we are partnering closely with data center customers to support and where possible, accelerate their qualification time line on our high-capacity Mozaic products. As Dave highlighted earlier, a majority of the largest cloud customers in the world are now qualified on our HAMR-based Mozaic drives, and we are continuing to ramp these products to support customer demand. The strong data center growth that I just described more than offset lower sequential sales in the edge IoT market, which made up the remaining 20% of revenue at $515 million. We are expecting some seasonal improvement in Edge IoT revenue in the December quarter from both VIA, Edge and consumer products. Moving on to the rest of the income statement. Non-GAAP gross profit increased to $1.1 billion, up 14% quarter-over-quarter and 46% compared with the prior year period. We expanded non-GAAP gross margin to 40.1%, which represents an incremental margin of nearly 70%. This margin growth reflects the benefit of increased adoption of our latest generation products and ongoing execution of our pricing strategy. Non-GAAP operating expenses were $291 million, up 2% quarter-over-quarter and in line with our expectations. The combination of strong top line growth and significant financial leverage drove a 19% improvement in operating profit to $763 million. Other income and expense were $74 million, and we are currently projecting OI&E to be essentially flat in the December quarter. We grew non-GAAP net income to $583 million with corresponding non-GAAP EPS of $2.61 per share based on tax expenses of $106 million and a diluted share count of approximately 223 million shares, including the net impact of our 2028 convertible notes of approximately 7 million shares. Turning now to cash flow and the balance sheet. We invested $105 million in capital expenditures for the September quarter or roughly 4% of revenue. For fiscal '26, we anticipate capital expenditures to be inside our target range of 4% to 6% of revenue, while we continue maintaining capital discipline. Free cash flow generation was flat quarter-over-quarter at $427 million, including the substantial variable compensation payout we discussed on our July earnings call. Looking ahead, we expect free cash flow generation to expand in the December quarter. We returned $153 million to shareholders through dividend. And as Dave noted earlier, we are increasing our quarterly dividend by approximately 3% to $0.74 per share. We deployed $29 million to repurchase shares of our common stock at an average price of $187 per share. We will continue to opportunistically repurchase shares and anticipate share repurchase activities to vary from quarter-to-quarter. We remain committed to returning at least 75% of free cash flow to shareholders over time. Cash and cash equivalents increased 25% sequentially to close the September quarter with ample liquidity of $2.4 billion, including our undrawn revolving credit facility of $1.3 billion. We exited the quarter with gross debt of approximately $5 billion and net leverage ratio of 1.5x based on adjusted EBITDA of $831 million for the September quarter, up 19% quarter-over-quarter and up 67% year-on-year. We are pleased that our strong execution is being recognized with S&P upgrading our credit rating earlier this month. Looking ahead, we expect net leverage ratio will continue to trend lower as profitability increases in the coming quarters. Additionally, we are exploring opportunities to further reduce debt, supporting the positive leverage ratio trajectory. Turning now to the December quarter outlook. The demand environment remains strong, particularly among global cloud data centers. We expect to increase revenue and expand margins as these customers continue to shift to our next-generation storage solutions to support their increasing demand. We expect December quarter revenue to be in a range of $2.7 billion, plus or minus $100 million, which represents a 16% year-over-year improvement at the midpoint. Non-GAAP operating expenses are expected to remain relatively flat at approximately $290 million. Based on the midpoint of our revenue guidance, non-GAAP operating margin is expected to expand to around 30%. Non-GAAP EPS is expected to be $2.75, plus or minus $0.20 based on a tax rate of about 16% and non-GAAP diluted share count of 227 million shares, including estimated dilution from our 2028 convertible notes of 10 million shares. As demonstrated by our September quarter results, Seagate is delivering on our financial commitments, reinforcing our track record of operational execution. Our performance is underpinned by a strong product road map that offer enterprise exabyte scale storage solutions, enabling them to maximize the potential of their data. This strength positions Seagate to drive meaningful value for both customers and shareholders. Operator, let's open the call up for questions.
Operator: [Operator Instructions] Our first question today is from Mark Newman with Bernstein.
Mark Newman: Congrats on a great quarter. The question was really, if you look at -- you commented at the beginning at the strong orders and it seems like you have order backlog through to -- I think you mentioned through to 2027. And the supply seems to be quite tight in the market. I just wonder if there is any plans to add capacity, if there's any specific supply chain bottlenecks that may alleviate over time? Obviously, there's pros and cons to that. So I'd just like to see how you're thinking about the whole supply-demand balance and if you're adding any capacity, that would be great if you could comment on that. And then a follow-up on HAMR mix. Congrats on the fifth hyperscaler being qualified, solid execution on HAMR so far. Just wondered what you're hearing from customers so far in terms of adoption for HAMR? Is there any upside or downside to the previous projections you provided for HAMR rollout going forward?
William Mosley: Thanks, Mark. Your two questions kind of go hand in glove. What I would say is that our strategy for adding capacity, if you will, is to go through product transitions. We're not really adding unit capacity. Through some of these product transitions, we actually lose a little bit of capacity because the process content is a little higher as we go through, but we add exabyte capacity. And to your question about HAMR, a lot of the reason that customers are engaging with us on these long-term agreements is because we have visibility into higher and higher capacity points. So the demand that most of the hyperscalers certainly are feeling is for more exabytes. More efficient exabytes in their data center allows their space and power and all their other metrics to get the best returns of TCO, if you will. And that's what we're really trying to answer the call for. We're very focused on going through the product transitions, getting our yields up on the product transitions and then ultimately transitioning to the point where we get 40s and 50s and so on and add exabyte capacity that way.
Mark Newman: Any update on the HAMR rollout, sorry?
William Mosley: Nothing more than what we said in the script, I think the qualification you made reference to. The ramp continues on and part of this is predictability of those transitions that we're going through. We have to make sure we're staged for that so that we get the customers what they need.
Mark Newman: Yes, we are happy to have achieved another call during the quarter. So now we have 5 customers qualified on 5 big customers, cloud customers qualified on HAMR. Of course, this is contributing to our delivery in the quarter and how we guided December quarter. So we are achieving those level of revenue and profitability a little bit faster than what we were thinking. And of course, this is related to the transition of the mix to higher capacity drives mainly to HAMR.
Operator: The next question is from Erik Woodring with Morgan Stanley.
Erik Woodring: And congrats on your results tonight. Gianluca, since your May Analyst Day, you've reported 2 quarters where your incremental margins have been 60% to 70%. Is it fair to say that in the new demand environment that we're in and the need for higher capacity drives, we should be thinking about your incremental margins just being higher, consistently higher than that 50% incremental margin you outlined at your Analyst Day? Or why would we not see these level of incremental margins sustain? I realize not literally every quarter. But generally speaking, why would we not see 60% to 70% incremental margins sustain from here?
Gianluca Romano: Thank you, Erik. Well, you're right. No, we are executing a little bit better than what we were expecting. And as I said before, this is mainly due to the move to better mix in terms of profitability. Not every quarter is the same. So of course, now that we are qualifying more customers and moving more customers to the HAMR drives. We have a little bit of a higher support in terms of profitability. I would say this will be different every quarter, even the pricing strategy that we have is very consistent, but not every quarter, we have the same number of new negotiations going on. So it's not easy to estimate exactly what will be the profitability and the mix of a specific quarter. I think in the short term, we are delivering very good results. Longer term, I think the model that we presented at the Analyst Day is a strong model. But you're right, in the short term, we are doing a little bit better. And you have seen how we got in December, it basically implies a higher margin than the 50% incremental from September.
Operator: The next question is from Jim Schneider with Goldman Sachs.
James Schneider: I was wondering if you could maybe address the level of cost reduction you expect to achieve on a blended basis as we look out into calendar 2026. Is there anything that would kind of prevent you from achieving that kind of mid-teens cost down on a blended basis, even as you ramp HAMR more aggressively, could that actually be better than that? Or is there a reason it would be worse than that? And then just to clarify, the prior statement you made on not adding unit capacity, is there any kind of benefit you get in terms of sort of dual capacity tracking HAMR versus conventional technologies that would add a little bit of unit capacity into the overall mix? Or is that just an overall dilutive event to overall units?
William Mosley: I'll take the second part first, and then I'll let Gianluca answer the cost question. So the way I think about it is, and I think I understood your question this way, there's some of our PMR technologies, which are in high volume, and as we transition to HAMR and the newer products, 30, 40 terabytes [indiscernible] then we're going to have to pivot pretty substantially. We don't really think about it as going back and building more of the PMR technology. We think about it as freeing up the PMR technology to ultimately go through the pivot, and that's how we're going to add exabyte capacity. As far as unit capacity, that -- a lot of that comes down to customers that are asking or qualified and what we've planned with them. So those plans are taking a long time to develop. And what Gianluca said earlier, if we get any upside -- marginal upside, it's because we're pulling in the future a little bit. Do you want to answer the question on cost, Gianluca?
Gianluca Romano: Yes. On the cost, of course, we don't guide cost for calendar '26. I would say the good improvement in our cost per terabyte is coming from the transition to higher capacity drives. So the mix is very important and every quarter is different. I would say you will see more and more transition to HAMR. We have 2 customers that are qualifying 40-terabyte drive. That , for sure, will be a good boost to our reduction of cost per terabyte during calendar '26.
Operator: The next question is from Asiya Merchant with Citi.
Asiya Merchant: There was commentary earlier about inference demand. If you could talk a little bit about the visibility of that demand. What gives you confidence that this one continues to grow maybe in terms of applications that you're seeing and that's kind of giving you the confidence of demand through calendar year '27. And how should we think about seasonality here? Typically, March does have a seasonal drop? How should we think about that given AI inferencing demand is pretty strong here?
William Mosley: Yes. It's actually very interesting for all of us to watch. Most of what I think has driven our demand over the last couple of years is this move to video where short-form or long-form video, like over 80% of the Internet traffic right now is on -- is video content. And there's just literally tens to hundreds of millions of videos being uploaded every day. I think exactly to your question, the more people can generate, access those videos via inferencing, the faster that happens, the better it probably is for our storage. And so do we see -- how is that going to progress in the next 9 months or 12 months? It's a little hard to tell, but we're pretty excited as we see new applications coming online that allow people to generate the videos. The videos are longer in format. They're richer in content. They've got more embedded video, so on and so forth. But it is really hard to predict. And I think our customers are struggling with this a little bit as well, which is one of the reasons the demand is strong.
Gianluca Romano: Yes. On the seasonality in the March quarter, of course, it's a bit early to discuss about March. We just guided December. I would say considering that data center revenue is 80% of our total revenue, the impact of seasonality is probably lower than what you have seen in the past. I think every year will be a little bit lower. But of course, we are not guiding much, but we expect a good quarter.
Operator: The next question is from Wamsi Mohan with Bank of America.
Wamsi Mohan: Can you talk a little bit about how you're managing pricing in this very constrained environment? How much is contractually locked in going into a quarter? How much flexibility do you have in intra-quarter basis? And when we look at just the reported quarter, right, like the dollars per terabyte decline was consistent with the prior quarter. But obviously, the demand environment is very strong. So hoping you could unpack that a little bit. Is there a differential there between HAMR and non-HAMR that's contributing to that? Or what's causing that dollar per terabyte to be -- decline to be relatively consistent, given just like this very strong demand backdrop?
William Mosley: Yes. Thanks, Wamsi. So there's -- a lot of what we're doing right now is very predictable. To our earlier comments, if we can execute a little bit better than planned, we can take cost out or we can pull in products that get qualified faster, but we are supply limited from that perspective. And contractually, we are giving our customers that predictable economics as well. Over time, if we go through these transitions a little bit faster, then the next contract comes up and the next contract comes up, we can actually make sure we get the right economic returns for what the market needs. And I think that's the long-term strategy right now. But kind of near term, everything is happening according to plan to being slightly shifted left.
Gianluca Romano: Yes, our pricing strategy is the same since about 10 consecutive quarters. So when we renegotiate a contract, we increased -- slightly increased pricing for the same product. And then when customers move to higher capacity products, they can get a little bit of a lower price per terabyte. That's why with major transition of customers to HAMR product with higher capacity, you can see a slight decrease in the average price per terabyte. But you see in the profitability the impact of the like-for-like price increase and all the cost per terabyte decrease due to the mix.
Operator: The next question is from C.J. Muse with Cantor Fitzgerald.
Christopher Muse: I want to clarify the March seasonality question earlier. Curious, if you were to assume that consumer were seasonal, could you pivot supply more to the cloud? And then I guess as the main question, your customers are turning to SSDs, given the tremendous tightness on the HDD side. Curious your thoughts around that cannibalization. And I guess what would maybe change your mind in terms of the vision for this sustainably higher demand and then potentially add capacity to support that.
William Mosley: Yes. Thanks, C.J. So I don't think that customers are really changing their architectures because of what they're seeing right now. What everybody is driving us to do is getting more predictable over time and if anything, be more aggressive on the product transition. So I don't really think there's any "cannibalization. " As a matter of fact, I don't think it's in anybody's economic benefit to do so. And the architectures are pretty well set going out for a couple of years.
Gianluca Romano: Yes. No, we see actually demand -- the gap between supply and demand getting a little bit bigger every quarter, that means demand is shifting more into the future, is not taken by any other technology.
William Mosley: And then your comment on seasonality is kind of interesting to think about the -- in some of the edge IoT markets, there is seasonality indeed. We have been taking slowly. We've been taking some supply out of Edge IoT products and putting it into cloud as we can pivot demand. Some of that's happening naturally as a part of these product transitions, which we've been talking about. It's not something we can do very quickly until we get to some of the products like the 4 terabyte per platter where we have commonality all the way through the platforms, but we will think that. So even though the edge IoT is -- the revenue is going down a little bit, the profitability is actually greater because -- those products are being prioritized a totally different way than there was a ton of supply on the market. I do think that will mean a little bit muted seasonality because that market is still fairly strong. I think it was over $0.5 billion last quarter on the Edge IoT.
Operator: The next question is Krish Sankar with TD Cowen.
Sreekrishnan Sankarnarayanan: I had a question and a clarification. Dave or Gianluca, in the past, you spoke about sharing the cost benefit of HAMR with your customers. But now with cloud becoming a bigger portion in the AI tailwind, I'm wondering if you're rethinking your pricing strategy for HAMR i.e., can you increase it further? And then a clarification, Gianluca, historically, your revenue guide had a range of $150 million, now it's more like $100 million. So is that because better visibility build-to-order helping you tighten that range?
William Mosley: Yes. Thanks, Chris. So I think what you said is very true. We are -- as we go through these product transitions, we're being as predictable as we can with the customers. But they -- some of the contracts are quite long as well. And so the market -- as the demand goes up, the market will adjust. Remember, that most of the benefit that the customers are seeing is in that TCO proposition with a higher capacity drives. So the pricing is a factor in that, but also this benefit that they see in their TCO is a factor as well. And so it will adjust very slowly over time as the market develops.
Gianluca Romano: On the HAMR pricing, in the past, we discussed only with 1 customer to giving them a slightly lower price, but this customer, of course, help us with the first qualification of the product. And so for a certain volume, they have a lower price than other customers. But of course, this is transitioning fairly quickly. It's just a matter of a few more quarters.
Operator: The next question is from Amit Daryanani with Evercore ISI.
Amit Daryanani: I guess, Dave, as you see an uptick in video creation, you've been talking about this, but with offerings like OpenAI, Sora. And given your nearline capacity is fairly committed through calendar '26, are you seeing customers looking to potentially fund or co-invest CapEx dollars for Seagate to get access to more units? And is that something you'd be open to? I would love to just understand like if this sort of keeps playing out the way you outlined, is areal density going to be enough? Or would you or your customers have to eventually add some capacity to get more units? I'd love to understand if you'd be open to that co-investing angle. And then if you just qualify this a little bit. When you talked about shipping 1 million-plus Mosaic drives this quarter, does that imply that about 36 exabytes of the total units were HAMR-driven this quarter? Is that fair?
William Mosley: I'll let Gianluca answer the question on exabyte because he'll go calculate it, but you're not far off. I mean, what we -- the way we think about it is going through those transitions opens up all these higher capacity points. Yes, from a customer perspective, I think I mentioned earlier, there isn't great visibility on what some of these new tools are going to unlock. But I think there's a lot of optimism around demand, especially in the video properties that exist in the world. And so therefore, nobody wants to be too late to these. I don't think that there's any significant changes to architecture. But I do think what is -- what we see driving or what's driving us is really stability. So it's customers showing up and saying, let's be very predictable as far as what I'm going to get. Some of the numbers about how much demand there is above and beyond what our supply is. Those numbers are contacted by someone else. I don't think it's -- that's the way necessarily each customer looks at it. What customers are starting to see is a temporal shift. So saying, hey, can I pull this in a little bit because we know -- they know we're gaining exabyte capacity over time, especially through the product transitions. That's what you saw contribute a little bit more to last quarter. And as we go as hard as we possibly can through the product transitions, we'll be bringing on more and more exabytes. So getting these qualifications done, getting us up to ramp, it's all priority for the customers as well. That's the way they're going to add capacity.
Gianluca Romano: Yes. No. First of all, I think it's very important that we keep the current balance between supply and demand and not taking action to oversupply in the future, the industry. In term of exabyte, it's a nice question. I would say you know that our HAMR product is between 30 and 36 exabyte -- sorry, terabyte per unit. So with 1 million units sold in the quarter, we are into that range.
Operator: The next question is from Aaron Rakers with Wells Fargo.
Aaron Rakers: Yes, I guess, first on the housekeeping side. I know that new disclosures. I'm curious of how -- or if there's any way we should think about the systems business within -- I guess, it's in the data center piece of it that we kind of just think about what it was over the past many quarters and kind of think about that kind of level going forward? And then on a nearline demand perspective, given the growth that we're seeing, looking back at the Analyst Day, I know you outlined kind of a mid-20% CAGR. Has your thoughts at all changed whether or not that, that's structurally just looking forward just higher relative to what you initially thought back a few months ago?
William Mosley: Thanks, Aaron. From the systems business, we -- obviously, we package the drives into the racks to save some customers some work. It's a fairly limited number of customers. They're great customers, and we don't really see the landscape changing too much on that. There's -- the customers that are buying their drives today and having someone else do the integration aren't necessarily pivoting the systems could happen. But -- so I think [ study ] goes on that front relative to nearline, we're all watching demand. And the mid-20 number is something we continue to grapple with, and thus most of the questions we've been asked today about how much supply is there. Again, I'll just say it the way we bring them more exabyte supply is to get the bigger drives out. And so that's what we're all focused on.
Gianluca Romano: Yes, Aaron. And we report system as part of data center, and this is very similar to what the rest of the industry is doing. So we try to align, so it's easier to look at the different players in the same way.
Operator: The next question is from Thomas O'Malley with Barclays.
Thomas O'Malley: I wanted to understand the timing of the crossover between the Mosaic 3 platform and the Mosaic 4 platform. I think you guys have historically talked about the first generation is what takes a long time. That's what was the struggle with your first big customer, but now you're really seeing that adoption kind of accelerate across other global CSPs. So at Mosaic 4, it sounds like it's starting to ramp in volume in the second half fiscal year '26. Could you get to a crossover point in the first half '27? Like should we be thinking about 15% or 20% contribution? I'm just trying to understand how much supply you can actually add to the industry with that transition in the short period of time.
William Mosley: Thanks, Tom. Yes, it's a question we're asking as well because as we go up the ramp, then yields and scrap and everything else on the new product is what dictates that. We understand the old product pretty well. There is a lot of commonality and piece parts between the 2 technologies between the 2. So it's not -- this is not a complete unknown on the ramp. But from my perspective, it will be a little bit faster ramp. We have to go execute, and that's what the team is going to be focused on. The qualifications are running well so far. So there's no reason to be disappointed at this point, and I'll be continuing to employ the teams to get the yields up as quickly as we possibly can to accelerate the transition.
Operator: The next question is from Karl Ackerman with BNP Paribas.
Karl Ackerman: Dave, you spoke about longer-term agreements offering visibility into 2027. Presumably, that visibility is on exabytes for HAMR drives. But since you didn't specify what that visibility is, does exabyte slowdown in the context of that 25% exabyte growth over time? Or do you think expect demand actually accelerates into '27 from what you see today?
William Mosley: Well, I think that's a -- so there's a supply question and a demand question. I think, yes, the supply would go up quite a bit because as we go transitioning to the earlier question Tom asked as we transition to the 4+ product family, we get a lot more exabytes out of it. So this will drive that as hard as we possibly can. We are working with customers, major hyperscalers and everything else on the 5 that we've qualified on 3+. Now we get as many qualified on 4+ over the course of the next year, and we get significantly more exabytes out. That's what we're driving.
Operator: The next question is from Timothy Arcuri with UBS.
Timothy Arcuri: I also had a question about the HAMR crossover. So 5 of the CSPs are qualified now and the other 3 are going to get qualified in the first half of next year. It sounds like it's about high teens of exabytes now, maybe pushing 20%. I would have thought you could get to exabyte crossover before a year from now if more than half the CSPs are called now and everyone is pushing to get these higher cap drives. So why would it take another year to go from 15% to 20% of exabytes to more than half of exabytes?
William Mosley: Yes. Thanks, Tim. So there's a lag in the supply chain, obviously. We have -- we've started wafers for various products. We will sell those various products. Just pulling in the qualification does not necessarily mean we turn a light switch from the old product to the new product. We have to actually go through that transition ourselves in our own factories. We're going to do that as aggressively as we can. And I think I asked this back -- I answered this back in Tom's question, there is some commonality between. So we can pivot a little bit, but not as hard as we'd all like. And I think some of this will be dictated by our yield ramp and so on the new products. But things are going well. So we're going to be as aggressive as we can.
Gianluca Romano: It takes time to ramp. Now as you know, the cycle time for HAMR is not short. So we need to qualify first and then to ramp the product. So it takes a certain number of quarters to really go up in capacity.
Timothy Arcuri: So I guess that's why you'd be buying heads from TDK basically from the outside, right?
William Mosley: No, we're not buying heads today from TDK. And I would say TDK is a great partner, has been for a long time. We talked to them about technology all the time, but they don't have HAMR technology. So what we're focused on is how do we get transitioned to the HAMR exabytes as quickly as possible.
Timothy Arcuri: Yes. I meant PMO heads, but thanks.
Operator: The next question is from Steven Fox with Fox Advisors.
Steven Fox: Just listening to all the questions and thinking back to the presentation in May where you lined up sort of the road map and the time it takes. It seems like hard disk drives is going to become more of a bottleneck to some of the expansion plans for the cloud guys as we get into next year. How legitimate is that of a concern? And what else could be done at the cloud guys to just sort of leverage existing capacity to sort of get through that period and keep you guys on track?
William Mosley: Yes, Steve, I don't look at it as something that's immediate or going to be solved in a period of 3 months, the industry is not going to bring on more supply in 3 months. It's not -- so really what's coming out of all of this is customers are getting very predictable on long-term plans. And that's what they need to do, and that's that we need them to do as well because we've got long cycle times as we've been saying. So it's building industry health and that's good. We can't react with supply to everyone's wish list, and there's probably wish list that aren't real at some point. But the customers that we're working with are being very predictable for us telling us exactly what they're going to need, and we're answering, call it, to the extent that we can. And if they need more exabytes usually, we're off book. It's a timing problem, a temporal problem. We're up by a few months or 6 months or something like that. And so ultimately, as we go through these transitions, we're going to bring on more exabyte capacity and they're going to be happy with it. Again, we don't see any evidence architectures changing or anything like that.
Steven Fox: Understood. And the advanced algebra on how you're upsiding is just different every single quarter depending on -- you mentioned like half a dozen variables between how many exabytes you got out this quarter versus last quarter and the next quarter?
William Mosley: Yes. We're executing well, as we said in the script, and we're pulling as hard as we possibly can, but we're executing the plan that we have and pulling in as much as we can as qualifications complete, and we'll continue to do so to the extent that our factories will let us. I mean we're -- it's a fairly long supply chain.
Operator: The next question is from Ananda Baruah with Loop Capital.
Ananda Baruah: Dave, how should we think about -- well, what's the most useful way to think about velocity, I guess, the pace at which customers going forward will look to mix up to higher capacity points given the shortages and given the areal density is what's going to bring new exabytes to market, do you think that we'll see folks look to mix up faster than historical? And part of this, I get is obvious because when you make a switch to HAMR, there's like a stair step-up in capacity. But even as HAMR normalizes, let's say, once you get past crossover, say, 12 months from now, once HAMR normalizes, do you think we could see an even faster than typical mix up inside of HAMR? And if not, like what would be the gating factors to that? Appreciate it.
William Mosley: Thanks, Ananda. It doesn't seem like that long ago that we were talking about 16s going to 18s or 18s going to 20s, diminishing returns, I would say, on PMR products. And then as we've gone 30 to 40, we see an immense pull for 40s. And then -- and I think that will be true with 50s as well as we try to drive that part of the transition because the TCO proposition is so much better if you're building a data center and then you want those products. So I do think that, that's reflective in the customers' behaviors exactly to your point. Back in the day when we were moving from 16 to 18 or 18 to 20, we just weren't seeing that much push. Now on the demand side, on the true end demand side, I think it's also because of what's going on in the world, we pointed out video in the script. Some of the video properties are just exploding right now and the capability to create and diversify the video in the world is great, and people are monetizing it, that's great as well. So human creativity is what's fundamentally driving all of this stuff. And I don't see it slowing down.
Operator: The next question is from Tristan Gerra with Baird.
Tristan Gerra: Could you elaborate on the duration of the long-term agreements for HAMR? Is there a pricing component to it? And what percentage of total revenue is currently based on those agreements? How does that compare with what the mix of those agreements was last year?
Gianluca Romano: Well, every customer has a different [ FTA ] or build to order. So the duration is different, the volume is different, the mix is different. But what we said in the prepared remarks, the vast majority of our nearline exabyte have already been committed for entire calendar '26. So we have a fairly long build to order in place.
Tristan Gerra: Okay. And then as you look at the ramp of HAMR and you're not ramping capacity, you're migrating to higher densities. Is your expectation on the basis of that ramp and what you see in terms of demand that lead times are going to remain similar to what they are today? Or would you expect lead times to expand further despite the ramp and migration to higher density HAMR?
William Mosley: Yes, it's a good question, Tristan. I think they'll remain like they are today. Getting through the HAMR transition, you have to add a significant amount of content to wafer, in particular, which is one of the longest lead times. But the next generation and the next generation after that, you don't have to go through that amount of transition again. So I do think it's a step up in lead times. We'll work it to get it back down as quickly as we can, but I think there's a substantial amount of process content. So we've gone through that step up, at least on those products. And then after that, we don't have to get to 4 terabytes per platter, 5 terabytes per platter and so on.
Operator: The next question is from Mark Miller with the Benchmark Company.
Mark Miller: Congratulations on another good quarter. I'm just wondering -- are your margins and yields on your HAMR drives at parity with the legacy PMR drives? If not, when you expect that to occur?
William Mosley: Yes, Mark, I would say we don't have ePMR. So we have -- our last generation PMR drive is a great drive. It's fantastic. It's got great yields. I'm very happy with it. We're always going to be working the next generation and the next generation as hard as we possibly can. The 4 terabyte per platter, which is what we're all focused on right now is where we need to get the yields up and it's still early in its lifetime. So we'll continue to put as much muscle as we have in Seagate to get that done this year.
Operator: The next question is from Vijay Rakesh with Mizuho.
Vijay Rakesh: So just I saw good numbers on the gross margin side, up pretty nicely sequentially. Just wondering with the drop-through on HAMR, would you expect to hit the mid-45%, 45% margins exiting fiscal '26? Or how should we look at that margin progression? And I have a follow-up.
Gianluca Romano: Yes. As I said before, from Erik's question, no, we are very pleased with the progression in gross margin and also with the increase in revenue. So we have achieved the $2.6 billion in revenue and the 40% gross margin a little bit earlier than what we were thinking. And we have guided December with an incremental margin that is higher than the 50% that we discussed in our financial model. Every quarter is a little bit different. Right now, we have a fairly quick transition from the PMR drive to HAMR drive. And with higher capacity drive, we get better profitability. So in the short term, we are progressing well. And as I said, every quarter will be a little bit different. I think the model that we presented in May is a strong model for the next 3 years, but it doesn't mean we cannot execute even better.
William Mosley: Right, Vijay. And as I reflect on a lot of the questions today, it's about how well do we don't know the demand further out and some of it's predictable, but the demand may keep on coming based on what we see in some of the end applications and then our ability to continue to work the yield issues around the new 4 terabyte per platter and get through that transition as quickly as we can, and that will help the cost side and supply side from an exabyte perspective. So that's what we're all focused on. I appreciate the question.
Vijay Rakesh: Got it. And Dave, just on the -- on your follow-up on the video side with the increased traffic, do you -- the longer-term exabyte growth that you guys had laid out on the 20%, 25%, do you see upside to that now given this significant increase in video traffic on storage?
William Mosley: Well, this is what we're all still studying, I think. We've seen, especially with some of the new AI generation -- AI content generation, which we talked about in the prepared remarks, we're studying how fast the uptake there is for some of these new creative capabilities that people have.
Operator: This concludes our question-and-answer session. I would like to turn the conference back over to management for any closing remarks.
William Mosley: Thanks, Gary, and thanks to everyone for joining us today. As you can see, fiscal 2026 is off to a great start. It's marked by strong operational execution and outstanding financial results. I want to once again express my gratitude to our employees, our suppliers, our customers and our shareholders for their contributions to Seagate's ongoing success. Together, we are advancing innovation to serve growing data storage demand and position Seagate for long-term value creation. Thanks for your continued support. We look forward to the significant opportunities ahead.
Operator: The conference has now concluded. Thank you for attending today's presentation. You may now disconnect.