Western Digital Corp (WDC) — Q1 2026 Earnings Call Transcript

Thank you, and good afternoon, everyone. Joining me today are Irving Tan, Western Digital's Chief Executive Officer; and Kris Sennesael, Western Digital's Chief Financial Officer. Before we begin, please note that today's discussion will contain forward-looking statements based on management's current assumptions and expectations which are subject to various risks and uncertainties. These forward-looking statements include expectations for our product portfolio, our business plans and performance, ongoing market trends and our future financial results. We assume no obligation to update these statements. Please refer to our most recent annual 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. In our prepared remarks, our comments will be related to non-GAAP results on a continuing operations basis, unless stated otherwise. Reconciliations between the non-GAAP and comparable GAAP financial measures are included in the press release and other materials that are being posted in the Investor Relations section of our website at investor.wdc.com. Lastly, I want to note that when we refer to we, us, our, or similar terms, we are referring only to Western Digital as a company and not speaking on behalf of the industry. With that, I will now turn the call over to Irving for introductory remarks. Irving?

Thanks, Ambrish. Good afternoon, everyone, and thank you for joining us today. Across industries, adoption of AI is expanding, fueling innovation, reshaping business models and ushering in a new wave of digital transformation marked by higher productivity and richer user experiences. As agentic AI begins to scale across several industries and multimodal LLMs become the norm, we are seeing a steady acceleration of AI use cases and applications, driving robust ongoing demand for the data infrastructure that enables this growth. AI is not only a consumer of data, but a prolific creator of data as well, both synthetic and real world. It is reshaping how data is being generated, scaled, stored and monetized. Data is the fuel that powers AI and it is HDDs that provide the most reliable, scalable and cost-effective data storage solution, playing a vital role in storing the ever-increasing zettabytes of data created by the AI-driven economy. To cite an example of how AI is transforming various industries, one of the world's leading medical institutions is using an AI workflow that analyzes over 7 billion images derived from 14 million deidentified patient records. This process enables predictive analysis, improves the speed and accuracy of diagnostics to deliver enhanced patient outcomes. Such applications are generating massive volumes of new data that is being stored. At Western Digital, we are also leveraging AI internally to enhance productivity and accelerate innovation across our organization. For example, in engineering, AI is helping to modernize our firmware, enabling us to deliver new features quickly to our customers and in a more cost-effective manner. In our factories, we are seeing productivity gains of up to 10% in select AI use cases. AI tools are improving yield, detecting defect patterns through intelligent diagnostics and optimizing our test processes. In parallel, they are also being used to up-level our technician capabilities, enabling them to perform higher skilled tasks, accelerating issue diagnostics and troubleshooting. Across corporate functions, AI is streamlining workflows, making the organization more efficient every day. The rapid adoption of AI and data-driven workloads at hyperscalers is driving robust demand for our products and solutions. To fulfill the demand for more exabytes of storage, our customers are increasingly transitioning to higher capacity drives. Shipments of our latest ePMR products offering up to 26 terabytes CMR and 32-terabyte UltraSMR capacities continue to grow at an impressive pace, surpassing 2.2 million units in the September quarter. Our ability to reliably scale our ePMR technology and transition customers to higher capacity drives is one of several ways we support the growing demand for exabytes. We are also investing in head wafer and media technology and capacity to drive areal density higher. In addition, we're increasing our manufacturing throughput by leveraging automation, AI tools and enhancing our test capabilities. We recently inaugurated our system integration and test lab, a 25,600 square foot state-of-the-art facility in Rochester, Minnesota, to enable rapid adoption of our next-generation high-capacity drives. This lab provides dedicated test capabilities that mirror our hyperscale customers' production environments, enabling collaborative integrated product development with our customers, accelerating qualification cycles, thereby ultimately shortening time to market for our products and time to value for our customers. The AI-driven growth in data storage is accelerating demand for higher capacity drives, which comes with greater manufacturing complexity and longer production lead times. As a result, our customers are providing greater visibility into their long-term needs, which in turn strengthens our partnership and helps us to support their future growth requirements. Our top 7 customers have now provided purchase orders extending throughout the first half of the calendar year 2026. And 5 of them have provided purchase orders covering all of calendar year 2026. I'm also pleased to share that 1 of our largest hyperscale customers has signed an agreement covering all of calendar year 2027. These commitments underscore both the essential role of our products in the AI data economy and our customers' strong confidence in our product roadmap, including the transition to HAMR technology. We are making rapid progress in our HAMR development and are on track to start HAMR qualification for 1 hyperscale customer in the first half of calendar year 2026. And to expand the qualification process to up to 3 hyperscale customers throughout calendar year 2026. The key focus of our qualification efforts is to ensure the highest level of reliability, quality and scalable performance so that once qualification is complete, our customers have strong confidence in our HAMR products and can rapidly deploy them at scale. This positions us well for the ramp-up of volume production in the first half of calendar year 2027. In parallel, we will begin qualification of our next-generation ePMR drives in the first quarter of calendar year 2026, building on our industry-leading ePMR technology, a trusted, scalable and proven solution that our customers are very familiar with and that has been used reliably in their data centers. Together, our ePMR and HAMR technologies will enable high-capacity drives that meet the growing demand for exabytes from cloud and AI workloads. Our platforms business is also sharing in the upward momentum, driven by overall growth of on-prem and cloud storage, including AI and social media applications. We will continue to invest in this business as more opportunities unfold and continue to scale up. Innovation lies at the heart of what we do. We continue to expand our proven ePMR roadmap even further while bringing new technologies, including HAMR to market. In parallel, our engineering teams are focused on improving data throughput speed and bandwidth of our drives as well as power efficiency. Major progress is being made on all fronts. And we will keep all stakeholders, including customers and investors updated on any new developments. Let me now turn to our quarterly results and capital allocation updates. For the fiscal first quarter, Western Digital delivered revenue of $2.8 billion, non-GAAP gross margin of 43.9% and non-GAAP earnings per share of $1.78. Free cash flow for the quarter was $599 million. This quarter, yet again underscores our business' strong free cash flow generation. We remain confident in the long-term strength of the business and our balance sheet. As a result, this quarter, we significantly increased our share repurchases, and I'm pleased to announce that we will increase our dividend per share by 25% to $0.125 per share. Kris will discuss our capital allocation in more detail later. Looking ahead, we're excited about the opportunities AI continues to unlock for our business even as we navigate macroeconomic uncertainties. For the fiscal second quarter of 2026, we expect continued revenue growth driven by data center demand and improved profitability led by the adoption of higher capacity drives. Let me now turn the call over to Kris, who will discuss our fiscal first quarter results and the outlook for the second fiscal quarter in more detail.

Thank you, Irving, and good afternoon, everyone. As a strategically focused hard disk drive company, Western Digital plays a critical role in enabling the data-driven AI economy. The company is executing well, fulfilling customers' rapidly growing exabyte demand while delivering strong financial performance. During the first quarter of fiscal 2026, revenue was $2.8 billion, up 27% year-over-year, driven by strong demand for our nearline drives. Earnings per share was $1.78. Both revenue and EPS were above the high end of the guidance range. We delivered 204 exabytes to our customers, up 23% year-over-year. This includes 2.2 million drives of our latest generation ePMR with capacity points up to 26 terabytes CMR and 32 terabyte UltraSMR. Cloud represented 89% of total revenue at $2.5 billion, up 31% year-over-year, driven by strong demand for our higher capacity nearline product portfolio. Client represented 5% of total revenue at $146 million, up 5% year-over-year. Consumer represented 6% of revenue at $162 million, down 1% year-over-year. Gross margin for the fiscal first quarter was 43.9%. Gross margin improved 660 basis points year-over-year and 260 basis points sequentially. The improved gross margin performance reflects continuous mix shift towards higher capacity drives and tight cost control in our manufacturing sites and throughout the supply chain. Operating expenses were $381 million, slightly exceeding our guidance range driven by higher variable compensation on stronger-than-expected results. Operating income was $856 million, translating into an operating margin of 30.4%. Interest and other expenses were $44 million, and taking into account an effective tax rate of 17% and a diluted share count of 369 million shares, EPS was $1.78. Turning to the balance sheet. At the end of our fiscal first quarter, cash and cash equivalents were $2 billion, and total liquidity was $3.3 billion, including the undrawn revolver capacity. Debt outstanding was $4.7 billion, translating into a net debt position of $2.7 billion and a net leverage EBITDA ratio of just below 1 turn. Operating cash flow for the fiscal first quarter was $672 million, and capital expenditures were $73 million, resulting in strong free cash flow generation of $599 million for the quarter despite the fact that we made our final repatriation tax payment during the quarter of $331 million. During the quarter, we increased our share repurchases to approximately 6.4 million shares of common stock for a total of $553 million and made $39 million of dividend payments. Since the launch of our capital return program in the fourth quarter of fiscal 2025, we have returned a total of $785 million to our shareholders by way of share repurchases and dividend payments. Also, today we announced that our Board has approved a quarterly cash dividend of $0.125 per share of the company's common stock, payable on December 18, 2025, to shareholders of record as of December 4, 2025. This marks a 25% increase over the dividend announced in April and speaks to the long-term confidence we have in our business. I will now turn to the outlook for the second quarter of fiscal 2026. This outlook includes our current estimate of all anticipated or known tariff-related impacts on our business in this period. We anticipate revenue to be $2.9 billion, plus/minus $100 million. At midpoint, this reflects a growth of approximately 20% year-over-year. Gross margin is expected to be between 44% and 45%. We expect operating expenses to decrease on a sequential basis to a range of $365 million to $375 million. Interest and other expenses are anticipated to be approximately $50 million. The tax rate is expected to be approximately 17%. As a result, we expect diluted earnings per share to be $1.88 plus/minus $0.15 based on a non-GAAP diluted share count of approximately 375 million shares. In closing, this was another strong quarter for Western Digital with results exceeding expectations. The guidance for next quarter reflects continued tailwinds in our business as we remain focused on strong free cash flow generation and demonstrating our commitment to creating long-term value for our shareholders. With that, I will now turn the call back to Irving.

Thanks, Kris. Our leading technology roadmap, combined with our scalable, reliable and strong product portfolio is highly recognized by our customers. This is demonstrated by the longer duration agreements we've signed with our major customers. Western Digital's consistent execution, combined with powerful AI-driven tailwinds, positions us to deliver strong results and robust cash flow over the long term. As data creation continues to accelerate, our innovation and operational and fiscal discipline enables us to capture these opportunities efficiently and drive sustained shareholder value. With that, let's now begin the Q&A. Ambrish?

Thanks, Irving. Operator, you can now open the line to questions, please. To ensure that we hear from as many analysts as possible, please ask one question at a time. After we respond, we will give you an opportunity to ask one follow-up question. Operator?

Storage demand is off the charts. And part of the great narrative for the HDD industry is an oligopoly acting very rational with supply. On the other hand, we're seeing SSD adoption rise for certain AI workloads given the tight overall storage supply. So my question, how do you plan to meet rising customer demand while keeping supply/demand in balance?

C.J., thank you for the question. I hope all is well. Our focus is really on a couple of things. One, ensuring that we continue to quickly and reliably deliver increasingly higher capacity drives. A good example is the current PMR product that we have, where we shipped over 2.2 million units last quarter, which equates to roughly about 70 exabytes of data in total. That product is expected to ship well north of 3 million units this quarter. So it’s a real demonstration of our ability to deliver exabytes to customers at scale. The second thing is that, as we've highlighted in the past, our unique innovation around UltraSMR. This quarter, our UltraSMR and CMR mix is roughly 50-50. As you recall, UltraSMR gives us a 20% capacity uplift over CMR and a 10% capacity uplift over standard SMR. Those capabilities, plus the fact that we'll be launching our next-generation ePMR drive very soon, starting qualification in Q1 of calendar '26, and we anticipate it will go into ramp in the second half of calendar year '26, will give customers an ability to take advantage of higher capacity drives. We’ve also been working very closely with customers to mix them up in terms of capacity points as well. If you go back a year, the average capacity for our top 7 hyperscale customers has increased 21% year-on-year. So that’s a very strong testimony to how capacity points in our drives have scaled up. We also continue to invest into areal density technology improvements and capacity as we stated from the very onset of us spinning out as the standalone hard drive company. Those investments will continue to be able to deliver greater areal density improvements without the need for any additional unit capacity. We're also looking at increasing our manufacturing throughput by leveraging more automation, AI tools that we highlighted in the script and also enhancing our test capabilities. This increase in productivity of our existing footprint will enable us to deliver more exabytes to our customers as well. And last but not least, as we highlighted in the script as well, the investments that we've made into our test labs to accelerate qualification is a key part of our ability to bring higher capacity drives faster to customers and therefore, fulfill the need for exabytes as well. And maybe just let me end my comments by being very clear about one statement: we are not adding any unit capacity to our portfolio right now.

C.J., do you have a follow-up?

Yes, Ambrish. I guess on gross margins, great, 660 bps uplift year-on-year. But obviously, we're always looking forward. So how should we think about incremental gross margins from here? Is there a framework that we should use?

Yes, C.J. So I'm really pleased with the gross margin in Q1, delivering 43.9% gross margin, which, as you pointed out, was up 660 basis points year-over-year and 260 basis points on a sequential basis. The incremental gross margin in the quarter on a sequential basis was approximately 75%. As you've seen in the prepared remarks, we've also guided for Q2 fiscal '26 with further gross margin improvement in the range of 44% to 45%. So that gives you 44.5% at the midpoint, which gives you on or about 65% of incremental gross margin on a sequential basis. Looking forward, obviously, as a company, we're going to continue to focus on further gross margin improvements. And I'm comfortable to have incremental gross margins on a sequential basis of approximately 50%, and that will drive some further gross margin improvement.

I think in the prepared remarks, you alluded to even further extending out UltraSMR. It's good to hear a kind of reaffirmation of the HAMR roadmap. But I'm curious if you could unpack that a little bit more if there's further room above and beyond the 36 terabytes that you see for UltraSMR; is there a 12-platter stack? I know one of your smaller competitors recently made some announcements around that. I'm just curious how far before we can get to HAMR if there's further potential upward expansion on average capacities?

Yes, Aaron. As we've highlighted in the prepared remarks, we've pulled in the qualification process of our next-generation ePMR product to the first quarter of calendar year 2026. Initially, in our roadmap, it was in the first half of calendar year 2026. In the current roadmap, the capacity points are scheduled to be at 28 terabytes CMR and 36 terabytes UltraSMR, but I'll say we have very innovative and creative engineers. They will obviously continue to push the capacity points, and we'll see where we get to by the time we actually get to production ramp and qualification completeness. On HAMR, as you mentioned, we also pulled forward the qualification process by half a year. As we've highlighted in our roadmap in the past, the plan was to start HAMR qualification in the second half of calendar year 2026. We've now pulled that in into the first half of calendar year 2026 with one customer, and we look to expand that to up to 3 customers by the end of the calendar year. That’s really a testimony to the comments I made last quarter, where I said I was very pleased with the progress that we've been making in terms of areal density improvements and our capability to build a highly scalable product. Our focus now is ensuring that we are able to deliver products with the right reliability and right yields that are similar in sort of capacity and capability to our ePMR portfolio, which is what our customers expect of us.

Do you have a follow-up?

Yes, I do. I guess thinking about sticking with C.J.'s comments, we're always looking forward. Historically, there's been some attributes of seasonality to think about in the March quarter, but it sounds to me like you're pretty much stocked out from a capacity perspective through calendar '26. So curious if you have any thoughts on how we should maybe think about seasonality or whether or not that even applies for the March quarter at this point?

Yes. We provide guidance one quarter at a time, but I want to point out that our business has fundamentally changed. Currently, about 90%, specifically 89%, of our operations are focused on data centers. This means there isn't much seasonality involved. Our performance is primarily influenced by the deployment schedules of our large hyperscale customers. Any seasonality is mostly related to the 10% to 15% of our business that comes from the channel and our client and consumer segments. Overall, it's fair to say that there isn't significant material seasonality in our business moving forward.

Irving, your February Analyst Day seems like it was a long time ago in the current market, even though it's only been 8 months. Back then, you discussed exabyte growth of around 16% to 23% and an annual price deflation per terabyte of about 7%. It appears that the market has changed since then. I would appreciate your updated insights on how we should consider the growth of these two metrics in the coming years. Any information you could provide would be helpful.

Yes. Thanks for the question, Erik. I think we gave a base case of 15% CAGR exabyte growth with an AI uplift case of 23%. We're definitely seeing exabyte growth trend more towards that 23% growth rate, especially as we get into these longer-term agreements. In fact, firm POs, we have pretty much throughout all of calendar year '26, and we have agreements now for '27 and discussions with customers for durations even longer than that. We are clearly seeing demand trending more towards that 23% range. And then on the cost side, I think the sort of mid- to high-single-digit cost down is probably still a safe assumption.

Do you have a follow-up, Erik?

I would like to hear your thoughts on how short you believe supply is in comparison to demand currently. Based on your timeline for new product introductions, when do you anticipate that supply will significantly increase, so that your electric vehicle growth will be more reflective of demand rather than supply constraints?

Yes. Thanks for the question, Erik. I believe in calendar year 2026, the supply-demand balance will remain very supply constrained as we ramp up new capabilities. We anticipate that more exabytes will likely start coming online in the second half of calendar year '27.

I guess maybe to start with, Irving, it sounds like you're pulling in, at least the start of the HAMR qualification a bit earlier than expected. Can you just talk about how long does it normally take for a product to go from qualification to deployment and do you see HAMR being roughly in line to that? Or could it be done quicker?

Yes. Thanks for the question, Amit. Yes, we are pulling in our HAMR qualification by half a year, as I mentioned, from the second half of 2016 into the first half of '26. If we use our ePMR portfolio as a proxy, we typically are able to go from start of qualification to completion and ramp in roughly 2 to 3 quarters. That's the sort of target that we're working to. And that's why we talked about the ramp in the first half of calendar year '27 for our HAMR products. But again, I reiterate our focus is really on ensuring that we not only qualify a product and can ramp it, but we're delivering a reliable product to our customers as well. The last thing we want to do is qualify a product, ramp it up, and then we have production-level challenges with our customers. So that's what our focus is on. But in the meantime, we serve our next generation of ePMR that we are starting qualification in calendar Q1 of '26. That we anticipate to qualify in 2 quarters and ramp very quickly thereafter as per the current generation of ePMR that we've delivered as well.

Do you have a follow-up, Amit?

I do. You folks talked about leveraging AI internally. Can you just talk about what sort of productivity savings you think Western Digital can realize as you deploy AI internally? And does that sort of imply that as revenues keep growing, even as you keep OpEx flat in the $370 million, $375 million range. I'd love to just understand what does AI implementation internally mean? What does that mean from a productivity or savings basis for the company?

Yes. Thanks for the question. We have a series of AI initiatives that spread across the enterprise, as we've highlighted in the prepared remarks as well. We are clearly seeing a benefit in our manufacturing operations with AI using cases where we're seeing 10% productivity gain. It's really resulting in faster, better yields and faster throughput of our products. We've also started to use AI in helping us rewrite some of our firmware. We are seeing gains in the space of about 20% productivity gains there. It's still early days, and I think there's quite a fair amount of experimentation and exploration, but we see tremendous opportunity in the sort of early use cases that we've been able to apply AI into the enterprise which has yielded very positive results.

This is Joseph Leeman on for Wamsi. How should we be thinking about the mix of a $2.2 million ePMR drive you shipped in the quarter? I'm not sure if I heard correctly; I think you said it was about 70 exabytes. So that's about 31 terabytes per drive. Does the mix change from quarter-to-quarter? Or is that just going to trend higher, especially once the next qualification comes through?

Yes. So your numbers are right. It was 2.2 million units sold and delivered, roughly 70 exabytes. This quarter, we are planning to ship over 3 million units. We don't anticipate the mix to really change that much. So it's pretty consistent based on the customer profile that we have.

Did you have a follow-up?

No follow-up.

I was hoping you could discuss the breadth and stickiness of the announced price increase you disseminated in September. In particular, since much of your volume is on long-term agreements, are ASP improvements only to volume that is not on LTAs? So could you talk about that? That would be helpful.

The letter we sent out, Karl, was mainly directed at our channel customers. It primarily impacts our client and consumer portfolio and likely affects the lower end of our nearline capacity drives, which is about 10% to 15% of our business. Our hyperscale customers with some purchase orders and long-term agreements are under separate commercial arrangements that were not influenced by that letter.

A follow-up for you, Karl.

Excuse me, if I may. It seems you have several months left to divest your remaining stake in SanDisk without facing a tax penalty. That investment in SanDisk is turning out to be quite insightful. Could you update us on whether you plan to divest the remaining stake and, if you do, what your plans for cash usage will be, whether for paying down debt, investing in head and media, share buybacks, or other purposes?

Yes. So during Q1 of fiscal '26, we did not monetize the remaining stake in SanDisk. We still have 7.5 million shares. It is our intention to monetize that stake prior to the expiration of the 1-year anniversary of the separation, which is February 21. Last time when we did the monetization, we did a debt for equity exchange and we haven't made up our mind how we are going to do it, but it could potentially be a similar transaction like we did the first time.

I wanted to go into the long-term agreements. Irving, during the pandemic, we've been conditioned with kind of the DRAM and NAND suppliers to think about long-term agreements to something that is really good while things are moving up and to the right and kind of get torn up when things correct. Could you talk about the hooks that are in these agreements? Are these take-or-pay? How are they structured so that you feel confident around your ability to get value for the length of agreements that you're signing?

Sure. As I highlighted for 5 of our hyperscale customers, we actually have firm POs. So these are not LTAs; these are firm POs that have been placed on us. And for 1 of our largest hyperscale customers, we have an agreement for all of calendar year '27 with quite a significant amount of commercial teeth in them. So it's quite a different environment where I would say we are moving to a world where we have firm purchase orders. And even with longer-term agreements, there are appropriate commercial terms in there to protect ourselves in the case of any adjustments in their forecast.

Do you have a follow-up, Tom?

Yes. I've been asking this question throughout earnings here. We heard from Lam about their impact on AI spend. I asked Seagate just on what they think on $100 billion of AI spend you would see from a benefit to their business. They kind of talked about a high single-digit percentage of CapEx traditionally has gone there. Do you guys have any different view or would you be more nuanced in the way you looked at that?

Yes, I would say it's a bit more nuanced. We do track it. There's not a direct correlation to it. Obviously, the big spend in AI goes to GPUs and HBMs and power. But we've seen the percentage of CapEx on HDDs go from probably low single digits to trending more towards the 4% to 5% range.

Congratulations on the strong execution. This year, it looks like nearline exabyte growth is trending more towards that sort of 35% range for the full year. You drove 36% year-over-year growth in June, 30% growth during the September quarter. You've got an order book that extends out over the next, call it, 12 months, which is reflective, like you said, of your customers' exabyte demand profiles. Does the forward exabyte demand profile really suggest the normalization back to a 23% demand CAGR as you talked about, Irving, or is that more of a supply constraint-driven profile and demand is really trending above that range? My point is that given all this AI infrastructure investment in compute, networking, memory and storage, a 23% bit demand CAGR may not be too conservative, but I wanted to get your views.

Yes, it's a good question. I would say it's still an evolving environment where the CAGRs continue to increase. As I mentioned, if you go back just less than 12 months ago, we thought mid-teens was the right number. We're now seeing trending to the 23% range. With potential as we fast forward to the '27, '28 timeframe to increase even more, but that's something we're working through customers to ensure that we continue to drive areal density improvements to be able to support the CAGR growth that they're expecting going forward. So it's something we're working very closely with them. I think the big difference is that in the environment that we're facing, we're getting much deeper insight into our customers' forward-looking exabyte requirements, a much closer partnership in terms of how they want to adopt higher capacity drives more rapidly to support their data storage requirements going forward.

Do you have a follow-up, Harlan?

Yes, just a quick follow-up. So on the UltraSMR mix shift, good to see the team at a 50-50 mix. I don't think you guys answered this question, but given the order book, POs, LTAs, what is the mix trend on UltraSMR into 2026? And is this mix shift towards UltraSMR a rather important part of the driver of the stronger incremental gross margin flow through?

Yes. We said we will see the mix of UltraSMR continue to increase over time, both as existing customers who have qualified UltraSMR increase their UltraSMR footprint and we have another 2 customers that are going through UltraSMR qualification as we speak right now. So we anticipate the take-up of UltraSMR to be an increasing part of our portfolio and continue to grow going forward.

Yes. And Harlan, just in general, the transition to higher capacity drives typically translates into a better gross margin profile.

And if I may add, Harlan, if I may add, this is Ambrish. Remember, UltraSMR is also translatable to our HAMR. So that's something to keep in mind as well.

Great results here. If I can, just trying to unpack pretty strong beat relative to the guide. Given that you guys are in these long-term agreements and there is capacity constraints, just if you could help me unpack what drove the upside? Was it some pricing that came through? Was there some extra drive that you were able to push through? I don't know if it was a mix shift. If you could just help me unpack that, that would be great.

Yes. So as it relates to the upside in revenue, it was mostly driven by great execution by our manufacturing operations organization, pushing really hard on the supply side and improving yields, improving throughput. That created some upside on the supply side for us from a revenue point of view. Also, on the gross margin side, we had some upside there, mostly driven by a strong price environment where we have seen some modest low single-digit ASP per terabyte increases on a sequential and a year-over-year basis. In addition to that, as I just indicated, the shift to higher capacity drives is definitely benefiting the gross margin profile. Our customers want more exabytes, and they know they can get more exabytes as they move faster to higher capacity drives. So that was definitely beneficial. And in addition to that, again, the operations team is executing strong on driving down cost internally throughout the supply chain. A combination of all of that provided some upside in the Q1 financial results.

Do you have a follow-up, Asiya?

Sure. And how should I think about then, given you guys have been running very well on your productivity initiatives, how should we think about that cost decline, especially given you have some calls that are ramping up faster than expected? How should we think about the cost declines here in the outer quarters?

Yes. Again, the team continues to execute really well. A combination of moving to higher capacity drives, which results in a lower cost per terabyte, but then also really working on productivity, yield improvements, test time reductions, and driving operational efficiencies throughout the whole supply chain. A combination of all of that is delivering the mid- to high-single-digit cost per terabyte reductions that we've indicated at the Analyst Day and that you have seen being executed in the last couple of quarters.

If I adjust your free cash flow for the tax payment, it's $930 million against the non-GAAP net income of $655 million. I'm assuming there's something unusually positive in that number. And I'm trying just to right-size how we should think about free cash flows relative to net income going forward because that's just a tremendous performance in 1 quarter.

Yes. So very pleased with the very strong free cash flow of $599 million. This is the second quarter in a row where the free cash flow margin is well above 20%. So great execution there. As it relates to Q1 of fiscal '26, we had a major reduction in our working capital. So in part driven by a reduction in our DSOs as the billings linearity during the quarter is very strong. Days of inventory were slightly up, but also the days payable went up. Great execution there by the team. Unfortunately, as you know, once you've obtained some major reductions in working capital, it's hard to repeat that each and every quarter. It's our goal to maintain it at this level, but you will not see the incremental benefit that we saw in Q1 of fiscal '26. Anyhow, I think going forward, I feel comfortable with a free cash flow margin in the plus 20% range.

Do you have a follow-up, Steve?

Yes. Just real quick on the prior question. So it's maybe a chicken and egg question, but you said the customers are recognizing the need to mix up to get the exabytes they need. So is it the fact that they're pushing harder on you that you're then pushing harder on your development team to get these higher mix products out? Is that sort of the dynamic that's going on?

I think it's a win-win scenario, Steve, that sort of both organizations are working very closely. Customers obviously want higher capacity drives to fulfill the exabyte demand. It’s also beneficial for them from a TCO standpoint. Don’t forget, when you have high capacity drives, rack densities are much higher, and therefore, TCO is much better as well. And from our standpoint, it's a great way for us to better support the demand that our customers have on us and for us to be able to support the strong growth trajectory that we are seeing both in cloud and in AI going forward.

Thank you, Steve. Operator, can we have the last question, please?

Strong quarter. This is Eddy for Krish. I do have a long-term question regarding the shortages. It seems like you and your main peer are very disciplined about adding capacity, which, of course, makes sense from a financial standpoint. But I do wonder how you balance that discipline on one hand with the risk of pushing customers more towards SSDs because they have no other choice, which in turn results in more NAND capacity in the industry, which lowers NAND prices longer term. So it's a tricky situation, and it would be great to know how your company is planning on navigating this?

Yes. Thanks for the question. It's something we look closely at as well. I think the good news is that AI, as we highlighted, is a prolific generator of data, and therefore, more data is getting stored as the value of data increases. So all boats are rising. The demand for NAND bits, hard drive bits and even tape bits are increasing as a result. There are specific use cases that make sense for them to use SSDs. But fundamentally, if you look at data center architectures and the tiering between SSDs, HDDs and tape that is unlikely to change over time, right? We anticipate that HDDs will continue to remain roughly about 80% of the bits that start within the data center. And it's also important to recognize there are some inherent TCO benefits of HDDs as there are reliability challenges in terms of the number of rights that QLC can handle as well. Given all that dynamics, we don't anticipate seeing any major change there. There may be quarter-to-quarter variations because of supply-demand dynamics. But sort of the 80% of exabytes being stored on HDD, we anticipate will be the case going forward as well.

Do you have a follow-up, Eddy?

Yes, sure. Thank you, Irving. Your main peer did purchase Intevac earlier this year, which sells equipment that is needed for HAMR. And you guys sounded pretty positive about the qualification. I do wonder if you have fully navigated the risk from the Intevac purchase or if it's something that's still in progress today.

Yes. We have fully mitigated the risks related to Intevac. As we highlighted when the acquisition first happened by our peer, all our HAMR development is actually being done on a separate system called ANELVA that's provided to us by Canon.

Thank you all again for joining us today and for your interest in Western Digital. At Western Digital, we continue to make good progress executing on our strategy. We look forward to sharing more with you on some of the exciting new innovations that we've been working on and the steps that we are taking to create long-term shareholder value. Let me close by giving a shout out to all our employees, our Western Digital drivers and our ecosystem partners who show up every day, making a difference for our customers, shareholders and each other. Thank you all very much, and have a wonderful day ahead.