Everpure, Inc. (P) Earnings Call Transcript
February 26, 2025
Earnings Call Speaker Segments
Thanks, and good afternoon, everyone. Welcome to the earnings conference call for Pure's fourth quarter of fiscal year 2025. Joining us are Charlie Giancarlo, our Chief Executive Officer; Kevan Krysler, our Chief Financial Officer; and Rob Lee, our Chief Technology Officer. Once Charlie and Kevan finish their prepared remarks, we'll open the line for questions. We issued our press release after the market closed, and it's available on our website, where this call is also being webcast live. You can download the slides for this webcast at investor.purestorage.com. Today we'll be making forward-looking statements, and those carry a variety of risks and uncertainties. Among them are statements about our financial outlook and operations, our strategy, technology and its advantages, our current and new product offerings, and competitive industry and economic trends. All forward-looking statements we make today rest on facts and assumptions as of today, and we have no obligation to update them. Our actual results could differ materially from what we forecast, and reported results shouldn't be viewed as a sign of future performance. Some of the risks and uncertainties tied to our business are discussed in our SEC filings, and we point you to those public filings. On this call, every financial metric and related growth rate is a non-GAAP measure except for revenue, remaining performance obligations, or RPO, and cash and investments. You'll find reconciliations to the most directly comparable GAAP measures in our earnings press release and slides. This call is being broadcast live on the Pure Storage Investor Relations website and is being recorded so it can be played back. An archive of the webcast will be posted on the IR website and remains the property of Pure Storage. Our quiet period for the first quarter of fiscal 2026 starts when business closes on Friday, April 18, 2025. With that, I'll hand things over to Charlie.
Thanks, Paul. Good afternoon, everyone, and welcome to the call covering Q4 and fiscal 2025 results. We turned in a strong Q4, beating guidance on both revenue and earnings; finishing the year also highlights the strides we've made, since our long-term data platform approach is reshaping the way enterprises view storing and managing data. During FY '25, we executed on a number of important strategic goals. To start, we secured a design win with one of the top four hyperscalers. That relationship is advancing nicely through the coming phases of testing as well as deployment planning. We keep making headway certifying our usage across several price performance layers. This hyperscaler will be able to run a uniform architecture across multiple performance tiers of its storage hierarchy thanks to Pure's DirectFlash technology. It additionally lets them shift substantial power and space toward new workloads while boosting their data storage performance, reliability, and lifetime. We also brought out our 150 terabyte DirectFlash module, which represents a fivefold capacity gain versus standard hard disks and 2.5 over the largest SSDs. A 300 terabyte DFM remains on schedule for delivery late this year. Beyond that, growing and expanding the E family over the year amounted to a major step forward in tackling the anticipated tenfold rise in unstructured data that AI will drive across the next 5 years. Compared with disk, this gives businesses far better economics, plus superior power and density efficiencies that will fuel disk displacement in data centers. Within existing hybrid and hard disk environments, the E family has turned into a formidable competitor. NAND prices climbing in 2024 versus disk competition did hurt gross margins during the year, something we anticipate easing in 2025. We're also still investing in our Evergreen//One program and seeing it succeed, drawing strong customer satisfaction and robust growth among existing customers. Arguably our biggest and most revolutionary technology leap last year came with the GA release of Fusion v.2. Offered as a non-disruptive upgrade to every pure customer, Fusion v.2 reshapes conventional enterprise data storage architectures, moving them from application data silos into enterprise data clouds; through Fusion, Pure is now able to give customers data set management services that support their global data estate; this is an enterprise area that mostly gets handled by hand, when it's handled at all. For enterprises around the world, manual management along with insufficient traceability of datasets ranks among the top causes of data sprawl and cyber risk. Fusion unlocks considerable new opportunity and value for Pure here. A familiar analogy is the best way for me to convey how transformational Fusion is. Ten years ago, lots of us relied on an external hard drive to add storage to our computers; it gave us extra storage and let us back up our files. But once it filled up, we had to swap it for a bigger unit and move the data off the old one onto the new. Imagine a coworker had a disk with unused room on it. Sharing it wasn't simple at all, and moving files between machines proved a hassle. And then suppose that external drive died — recovery options were basically nonexistent. These days, hardly anyone bothers with external drives. Everyone has shifted toward personal cloud storage offered by a wide range of providers. The advantages extend far past mere storage. Take unlimited storage scalability, for instance — when someone requires additional space, they simply pay for it, with no data migration needed. Their infrastructure gets consistent updates through non-disruptive software and hardware improvements. Put differently, the service never becomes outdated. Guaranteed resiliency means a user anticipates never losing their data. Global access lets someone reach their data from any location, on any machine, with search capabilities for locating content and simple file sharing via just a few mouse clicks. This approach has carried over into how coworkers operate within an organization. When a company provisions personal data storage through services like Google Drive or Microsoft OneDrive, it can apply its corporate data policy to employee data management — for example, controlling file sharing to satisfy corporate compliance requirements, managing file retention and deletion policies, enabling or disabling individual access when employees join, change role, or leave the company, and providing administrative global search for all stored files.
A large number of enterprise clients possess hundreds of storage arrays spread across the globe, and every one of those functions as a separate data silo. Pure Fusion v.2 represents an incredible leap forward, turning Pure Storage arrays from standalone units into what amounts to an enterprise data cloud. With Pure Fusion, every one of an organization's Pure arrays can be administered as a single unified system. Even better, the information kept on them can be handled as a data cloud instead of as separate silos. Fusion lets applications tap into the combined capacity and performance shared across every array, rather than being limited to whichever ones they happen to plug into directly. Fusion takes care of compliance policies automatically, ensuring data is handled securely and consistently. It strengthens cybersecurity and removes mistakes caused by human hands. It makes automated data set management possible, enforcing and monitoring company rules and policies. On top of that, it lets stored information be reached throughout the worldwide enterprise, tearing down data silos and delivering a genuine enterprise data cloud. Initial reactions from dozens of customers have been overwhelmingly enthusiastic. As businesses grapple with obstacles such as AI integration, cybersecurity dangers, and hybrid cloud optimization, the case for leaving behind conventional manual storage management is obvious. When it runs as a data cloud, Fusion handles data management automatically, boosts security, and makes hybrid on-prem and cloud environments simpler. It accommodates real-time AI access, removes the requirement for too many copies of data, and encourages API-driven innovation. Pure Fusion makes possible a basic transformation from data silos toward an enterprise data cloud approach to data management. Thanks to Fusion, businesses gain a storage solution that is accessible everywhere and can automate and streamline their data management across thousands of arrays as easily as if they were managing a single array. This lets an enterprise run all of its data storage on a workflow basis, exactly the way the Cloud works; this represents not merely an evolution, but a revolution in IT infrastructure, reshaping enterprise storage standards and data management. Several large enterprise deals closed successfully this quarter largely because of this broader platform strategy. Take, for instance, a financial services technology firm that rebuilt its whole storage environment with Cure after suffering slow response times under a legacy architecture. By bringing fragmented storage together onto our unified platform, this company now converts data into a powerful asset while handling hundreds of billions of financial transactions each year. A telco ranked among the biggest in the world picked Pure to address its serious data center power consumption problems and cut energy usage in data storage by an impressive 89%. Our 150 terabyte DFMs served as the cornerstone of that achievement, generating substantial business impact. Our solution let this provider handle petabytes of data, perform advanced analytics, and archive data over extended periods using the same core architecture. Portworx, our storage solution built for containerized applications in Kubernetes environments, experienced a strong rise during FY '25, fueled by considerable growth in AI analytics and Cloud-native platform use cases. In the last quarter, two Fortune 50 companies made the decision to move their VMware environments to a modern virtualization environment leveraging Portworx and Red Hat OpenShift. For one of these organizations, the pace at which applications get rolled out—and the time needed to see outcomes—will increase dramatically, and their expenses will drop by 30% as well. The second, a large-scale producer, is moving the containerized and virtualized application platforms at its manufacturing sites onto Kubernetes. With Portworx, that manufacturer can lower what it spends on running separate infrastructure for VMs and containers, and make operations less complicated. Our AI-related business keeps expanding, as I've mentioned during the past several quarters. Our high performance storage is being tapped more and more by machine learning and training environments, while AI data preparation chains increasingly rely on Portworx. Next month, at the NVIDIA GPU Technology Conference, we can't wait to show off our newest AI developments; we'll reveal how Flashblade establishes a fresh standard for performance, scalability, and deployment simplicity that nothing else matches in large scale AI infrastructure rollouts. Now for the broader economic picture: we anticipate that geopolitical unpredictability will keep conditions in flux, certainly for the remainder of this year. Beyond how we normally run things—with supply and distribution chains that are diversified and spread around the globe—we've put together contingency plans covering multiple tariff situations. All things considered, FY '25 was a performance we feel good about, and our confidence stays very high in the distinctive, leading data platform strategy and vision we have. Helping our customers construct their very own enterprise data clouds spanning their distributed and hybrid IT environments is something we're thrilled to do. The Pure platform's four key competitive advantages bolster this confidence. They are: our unified purity operating system, our evergreen technology and business model, our purity DirectFlash technology, and our cloud operating model—now improved by Pure Fusion v.2, letting our customers build their own enterprise data cloud; with that, I'll hand the microphone to Kevin.
Thanks, Charlie. We finished the year strongly, beating our guidance with revenue growth in the double digits, $153 million in operating profit, and an operating margin of 17.4%. During Q4, total revenue rose 11%, and TCV sales for Evergreen//One hit a record. In FY '25, we crossed a significant financial threshold by exceeding $3 billion in total revenue for the first time, with total revenue of $3.2 billion, up 12%; we also posted our highest annual operating profit at $559 million. Our Q4 performance was strong thanks to sales of our solutions across the data storage platform. Flashblade, FlashArray XL, Portworx, our E family, and renewals of Evergreen subscriptions throughout our install base all achieved record sales in Q4. A key strategic priority remains speeding up customers' shift of cost-sensitive workloads to our E family and FlashArray C solutions, as more and more customers see the better economics compared with traditional disk solutions. Record sales of our E Family solutions, together with higher QLC flash costs and relatively stable hard disk solution pricing, led to temporarily lower product gross margins of 62.9% in Q4. We also ended the year with record quarterly TCV sales for Evergreen//One of $140 million, up 20%. FY '25 TCV sales for Evergreen//One and our other service based offerings reached 393 million, a 3% decline. This stemmed from both longer timelines required to close Evergreen//One deals greater than 5 million and a higher conversion of Evergreen//One opportunities into a traditional sale that we saw last quarter; we remain confident about the long-term growth potential of our expanding as-a-service offerings. We expect TCV sales for Evergreen//One to increase next year, and we will keep providing quarterly updates, though specific guidance will not be given. Q4 subscription services revenue of $385 million was up 17%, and subscription services annual recurring revenue, or ARR, rose 21% to $1.7 billion. Total remaining obligations, or RPO, at the end of Q4, covering both subscription services and non-cancellable product orders, increased 14% to $2.6 billion. Apart from non-cancellable product orders, RPO tied solely to our subscription services rose 15%. The overall RPO increase was affected by the drop in TCV sales for our storage as a service offerings during FY '25. Our subscription services net dollar retention, or NDR, was 117% at year-end and remains in line with our long-term target of 115%. In Q4, U.S revenue of $619 million drove most of the growth, while international revenue came to $261 million, down 3% year-over-year. For FY '25, U.S revenue rose 12% and international revenue rose 13%. In addition, we added 334 new customers, roughly in line with both Q4 FY '24 and the prior quarter. We still serve 62% of the Fortune 500; in Q4, total gross margin was 69.2%, helped by a solid subscription services margin of 77.2%, while product gross margin, affected by record sales of our E family, was 62.9%. For the entire year, our total gross margin came in at 71.8%, versus 73.2% during FY '24. When we look forward to the full FY '26 year, product gross margins should land in the mid 60s, which falls within our long-term expectation range of 65% to 70% for product gross margins. Several things drive this outlook: anticipated robust demand for E family and FlashArray C solutions, plus our belief that QLC flash pricing will ease — a key consideration given that we compete with disk-based solutions. At year-end, our headcount was up modestly, reaching close to 6,000 employees. Our balance sheet stays strong, with $1.5 billion in cash and investments as of the year's close. Q4 cash flow from operations totaled $208 million, and for the full year it was $753 million. Capital expenditures for the year came to $227 million, roughly 7.2% of revenue for FY '25. The capital investments we made during the year backed data center expansion to support broader testing of new products and solutions — among them our recent hyperscale solution design win — along with increasing density of our DirectFlash modules and software development for our Fusion v.2 solution. Free cash flow was $152 million in Q4 and $526 million for the year. Our free cash flow margin for the year reached 16.6%. During Q4, we bought back 3.1 million shares, returning about $192 $million to shareholders. Across FY '25, share repurchases totaled $374 million, or 6.7 million shares. On top of that, we paid 65 million in withholding taxes on employee awards in Q4, offsetting dilution by roughly 1 million shares, and 209 million for FY '25, offsetting about 3.5 million shares. Right now, 21 million remains available under our existing repurchase authorization, and we're announcing a new share repurchase program worth an additional 250 million. Now to our FY '26 guidance: once again we anticipate double-digit revenue growth, rising 11% to slightly above $3.5 billion. Our annual revenue outlook assumes an IT spending environment comparable to FY '25, plus a return to growth for Evergreen//One and other service offerings; for Q1, we're expecting revenue of $770 million, which represents an 11% year-over-year increase. Turning to operating profit. We stay dedicated to our bold strategic initiatives, positioning the company for strong, sustainable, long-term growth; this involves leveraging our recent hyperscale design win through incremental investments to scale operations for large production and deployments beginning in FY '27, as well as accelerating density in our DirectFlash modules. We anticipate an operating margin of approximately 17% in FY '26, consistent with what I said last quarter, which translates to an operating profit of around $595 million. For Q1, we're guiding operating profit of $80 million and operating margin of 10.4%. To wrap up, we're pleased to have gotten back to double-digit revenue growth and to have exceeded $3 billion in revenue for the first time. This pivotal year has been defined by remarkable innovation and execution, laying the groundwork for sustainable long-term growth; with that, I'll hand it back to Paul for Q&A.
Appreciate that, Kevin. Ahead of opening things up for questions, please keep it to a single question with just 1 part, so that we're able to reach as many of you as we can. Should you have more questions, we'd ask that you kindly get back into the queue, and we'll gladly take those further questions if time permits. Operator, please go ahead and begin.
Great. Appreciate you guys fielding the questions. I'd like to revisit a few things from the prepared remarks regarding Fusion. I realize it's still early days since Fusion 2.0 became available, but are there any indications yet from customers who've taken up Fusion of either quicker expansions or greater stickiness? If it remains too soon, could you at least talk about the go-to-market side and where things stand when it comes to building adoption and awareness of what Fusion makes possible? Thank you.
Thanks, Mike. This is Charlie. Appreciate the question. So here's where things stand: before a customer can try out Fusion, they must grab the most recent update for the FlashArray software operating system. Dozens of businesses have done so, and in particular have switched Fusion on and begun using it. And with under a quarter of that having reached the market, like I said, dozens of customers have taken that step. That's a really quick pickup; and as I noted during my prepared remarks, naturally we're collecting feedback, and that feedback hasn't merely been favorable, it's been excited. They're discovering that being able to handle every system placed under Fusion as one fleet instead of separate arrays makes things far simpler; they're currently trying out what we refer to as presets. Presets let them establish policies governing how data gets treated throughout the arrays, so that for a given storage class, every bit of data within it receives identical handling. Thus instead of manually managing various application use cases when it comes to policy, the policy now gets executed through software. So the response has been extremely favorable. Clearly, this generates a network effect. It produces a network effect for our offerings, which means that should you have all, or perhaps a sizable share, of your Pure Storage arrays running under Fusion, the odds of you bringing in another array outside of it, I believe, will be quite small. That said, to your point, it's somewhat early right now. We're still in the initial stages of rolling this out. My expectation is that sort of network effect will show up once several quarters of use have passed.
Thanks, Mike. Let's take the next question.
Sure. I appreciate you fielding this. My question concerns the hyperscale opportunity; although I realize it's still early to fully dissect what was announced last quarter, I'm wondering whether there's any sense of how the model for this opportunity could be developing? And along those same lines, how might you describe the talks, or how extensive those talks are, that you could be holding with additional hyperscale opportunities? Thank you.
Yes. As you can probably guess, we're at an early stage, yet the dialogue keeps developing and honestly widening, when it comes to use cases for various kinds of data storage tiers within that hyperscaler; so we have been advancing and carrying on with greater depth in testing and joint development, especially, plus conversations about future scenarios and where that storage is headed. Compared with other vendors, we keep broadening our talks, and I'd say we told you previously that we anticipated they would speed up; through this quarter they indeed have sped up. There's nothing we can announce at this point, but we are definitely witnessing growing interest. Rob, would you like to say something?
No, I believe that covers it nicely. Things are moving along the way we anticipated, and like Charlie said, we're now advancing into the later stages — finishing up their implementation plans and going through the usual engineering work — which we'd anticipate would then bring about more thorough field testing, on the path toward scaling up in a big way. And yes, the talks and engagements we're having with additional hyperscalers are certainly picking up speed.
Thanks, Aaron. Let's take the next question.
Thank you very much. Good afternoon, everybody. Perhaps I'll just ask something that builds on what Aaron raised about hyperscale. Could you describe the investments Pure needs right now to take this from a design win and how it advances into a production environment? I'd really like to get a sense of the kinds of investments you have to make to reach that point? And then, if there's any clarity on the steps required to truly shift this from design today to, I believe, beginning production in fiscal '27, that would be useful to understand? Thanks.
Yes. So this is a continuing co-development effort with that hyperscaler, which means they're building new software, new architectures, new infrastructure architectures for, as we've talked about, their next-generation data center design. And while they're designing that software and those hardware architectures, we must work side by side with them to ensure our software operates closely with theirs. Naturally, we're always collaborating to try to enhance both performance and price performance in that setting. As we said, we're now working across multiple different performance tiers. Every one of those requires its own set of tuning. And naturally, we also must qualify several different scales of our DirectFlash modules together with several different Flash partners, so to speak, various flash manufacturers for those DirectFlash modules. Now, all of that demands a fair amount of work and testing on our end too.
Okay, Amit, this is Kevin. To pull together what Charlie was saying, I’d break it into four main areas. First, bringing the Purity software and our DirectFlash technology together with hyperscaler hardware specs. Second, speeding up our density roadmap through our DirectFlash modules. Third, getting more NAND suppliers qualified, which Charlie mentioned. And fourth, growing our supply chain capabilities and scale, plus investing to go after other opportunities, including hyperscaler opportunities too.
Thanks, Amit. Let's take the next question.
Thanks; this is for Charlie and/or Kevin — could you help us break down the product gross margin decline? How much stems from incremental sales of the E and C families, as opposed to, say, higher QLC costs or possibly more aggressive pricing in other parts of your portfolio? Thanks.
Yes. So, yes, gross margin performance in the other parts of our portfolio has remained fairly steady; so there's no real change in those areas; so the entire decline comes from the E family; now, that has two effects. First, the E family now represents a larger share of our overall mix. But the second effect is that the E family, since it goes up against disk, and in that space the rise in NAND prices and media prices didn't hit the disk market. So in that situation, while our margins have been very stable when we're up against other flash pricing, when we're up against disk solutions and NAND prices have gone up, that family feels it more, as you can imagine. So we see that as somewhat temporary, as NAND prices moderate in the coming year, when margins for the E family will return to what our longer term — or actually I shouldn't put it that way — what our early stage e margins were, because we do expect that as we gain market share in the E family, even the E family will begin to approach our normalized margins. But as we said many quarters ago, we intended to be aggressive with the E family given that we're really the only player in the market right now that can compete with disk, and we want to take advantage of that. So we are being aggressive there. The more recent uptick in NAND pricing last year affected us a bit more than one would have hoped.
Right, and on that note, Kevan here, Howard. To build on what Charlie said, when we look ahead to next year, we fully anticipate the momentum behind the E family carrying on. That said, given that QLC pricing is moderating, we do expect to return to a product gross margin range of roughly 65%, which is once again in line with the long-term profile for our product gross margin.
Thanks, Howard. Let's take the next question.
Hi everyone, I'm asking on behalf of Pinjalim. I appreciate you fielding our questions. Could you give us some detail on the geographic regions, since the U.S. beat international this quarter? Thanks.
Sure, Kevin here. Well, for one thing, we felt pretty good about how the U.S. performed, and the U.S. was the driver of that growth. Seeing quarter-to-quarter swings in growth across different regions isn't out of the ordinary. All things considered, growth ended up being fairly similar for the year, both across international markets overall and in the U.S. And our international revenue growth was really robust during the first three quarters of the year, then clearly declined in Q4, though there's nothing specific I'd point to, and over the longer term, we anticipate international markets and revenue expanding more quickly than the U.S., because of the scale of those markets and the opportunity there.
Appreciate it. Let's take the following question.
Hello, and thanks for fielding my inquiry. This one's directed at Charlie or Kevin. Regarding RPO, it appears the growth rate there has fallen — roughly 30% twelve months back, down to 14% during the most recent quarter. I'm wondering whether that's simply a function of scale, or if there's another factor at play with the RPN? And what are the implications for your main enterprise business? Thanks.
Right, Kevin here. I'll kick things off, then Charlie can chime in. Really, what we're seeing is a byproduct of the Evergreen//One TCV bookings we recorded this year. Naturally, with a slight revenue dip across the year, that weighs on our RPO for the year too — 15% once you strip out non-cancelable product orders. That said, we were very happy to see Q4 deliver a record Evergreen//One result at 20%. We've definitely factored growth into our revenue guidance for this year, Evergreen//One offering. And we're quite pleased and enthusiastic about that offering, along with the value our customers are realizing from Evergreen//One overall.
All right, that answers my questions, Kevin. I appreciate it.
All right, thanks, Krish. Let's take the following question.
Yes, thanks, and good afternoon, everyone. I'd like to build on what Rakers asked a bit earlier. Charlie, my sense is you didn't really address the matter of what the model looks like and what gross margin and operating margin could be for the hyperscaler business. Also, do you expect that you'll be in a position to offer more detailed guidance regarding revenue as the year progresses?
Yes, that's the answer there. Once we have a clearer picture of precisely how the hyperscalers intend to carry out their deployment plans — and we're assuming we ought to have that understanding somewhere in the 3 to 6 months ahead of the rollout itself — then we ought to be able to provide an update on precisely where things stand. What I'd say is that at present this is fairly dynamic; so until we're on somewhat firmer footing — dynamic meaning simply that we remain in the planning stage, with plenty of considerations around the precise timing — so the moment we reach a point where we have greater confidence regarding the precise timing and the rollout, we'll be in a position to provide further updates.
Also, Jason, one more point I'd make on the financial side: there really isn't anything fresh beyond what we went through at length during the prior quarter regarding that design win. When it comes to gross margin, once again, hardware wouldn't show up the way we're structuring it. Clearly, looking further out on operating margin, we anticipate expansion as time goes on thanks to this opportunity. That said, these elements line up closely with what we laid out at length three months ago.
Thanks, Jason. Could we take the next question, please?
Appreciate it. I'd like it if you could bring us up to speed on how you see enterprise uptake of AI projects potentially impacting your company. Charlie, I'm aware you've drawn attention to this uplift phase. Could you discuss your view of how that might influence your business, and the timing? Thanks.
Sure. As we've discussed previously, AI is a subject with many facets. For one, there's the large-scale machine learning setting, which is truly the top — the apex of the pyramid, so to speak, of companies that are going to run machine learning environments of their own. We did hint — I did hint — that a good new announcement is on the way that we'll be showing at GTC. That genuinely ties to the biggest-scale AI opportunities. The second facet is the inference opportunity. We've had quite a few discussions. We keep having discussions with customers and also with NVIDIA regarding the inference opportunity. In my view that's still somewhat distant, yet we're beginning to generate some early momentum through initial discussions with customers who intend to deploy in that space. Let me say again that in my opinion our greatest opportunity lies in changing how customers handle their production data, moving from data silos to an enterprise data cloud. We regard that as a major — we regard AI as a major prod, so to speak, toward customers reconsidering the architecture of their data storage so it becomes far friendlier, so to speak, to AI obtaining access to data for real-time analysis, particularly in inference and RAG type environments. So that's — we think Fusion couldn't arrive too soon, that it's a significant opportunity for us not just in the management space, in the compliance space, enabling customers to have a better non manual automated way of managing their data, but honestly in the AI space, making all of that data accessible for AI analysis.
Absolutely, Simon — Rob here, jumping in to build on what was just said. When we take a broader view, the AI category and what it offers us is, I believe, still unfolding very much the way we anticipated. As Charlie mentioned, there are those high-performance training infrastructure setups that have drawn considerable attention over roughly the past 12 months. What's beginning to fuel more of our customer engagements is that, as organizations shift further into inference or RAG deployment stages, or begin thinking through how they'll link their wider data landscapes, datasets, and silos into those AI workflows, they're turning to Pure — and to Pure Fusion — for assistance. So when we take a step back and consider our position within the AI category right now, FlashBlade has historically driven a great deal of that. Yet as the inference opportunity begins to unfold and, without question, as Fusion enters the picture, we're now witnessing a growing number of customers deploying our Portworx and FlashArray offerings — truly our full portfolio and platform of solutions — to assist them in advancing their AI efforts as they expand beyond a pure training focus into the full range of those environments.
Thanks, Simon. Let's take the next question.
Hi everyone, Quinton here filling in for Jim Fish. We appreciate you answering us. There are perhaps a pair of connected items regarding margins that we'd like to ask you, Kevin. In what way, if any, did you factor tariffs into the guidance you've given? Also, I believe I caught you mentioning that the money being spent this year is meant to enable additional hyperscaler possibilities. So is it fair for us to conclude that a possible new announcement during this year would not call for extra outlays, meaning the roughly 17% margins we're locking in are more or less protected? Thank you.
Yes, that works. Which one do you want to go with?
I'll take the hyperscaler question.
Yeah, so, what I'd say is we're expanding the ecosystem around the hyperscaler space. That involves scaling up our supply chain as a whole so it can handle both the higher volume coming from that single design win and additional wins too. Even so, that doesn't mean we won't need to invest in growing other opportunities. What I mean is, every opportunity on our radar comes with some level of customization, specifically around how our Purity software integrates with hardware. Still, it's clearly a massive long-term opportunity from where we sit. Charlie, would you like to chime in on tariffs too?
Yeah, there is somewhat of a tariff dimension to this. To start with, I believe we demonstrated during COVID, when supply chains were disrupted in extraordinary ways, that we are extremely agile, very capable, and that we have a supply chain that is both robust and dependable, with plenty of diversity. You might remember — or if you haven't been tracking us for quite that long, you may not know — that at the worst point we reached a lead time of 6 weeks, and that was a leading position not only within our own sector but throughout all of technology; since that time we have added even more diversity to our supply chain and to how we distribute. At this moment, I cannot forecast, and I doubt anyone else can either, precisely which tariffs will be imposed, on what schedule, or for what duration. For a wide range of possible situations, we are putting contingency plans together. Tariffs will be watched closely by us, and when the moment is right, we will do what is necessary and give an update once that happens.
Let's take the next question.
Thank you very much. Charlie and Kevan, you mentioned a moment ago that the E Series now accounts for a bigger portion of the mix and that you plan to push harder on the E family. So given that you're going after this opportunity, why isn't a stronger reacceleration in revenues being projected? I mean, your exit rate for this year sits at 11% growth. You're guiding to 11% growth, assuming nothing changes on the macro side, yet you're pursuing this opportunity very aggressively. So is something else offsetting it, like how should we think about the absence of reacceleration when it comes to revenue? Thank you.
Yes. When you take into account that we also expect growth to return in the Evergreen//One business. I think it's a we believe it's a sensible forecast or guide for the year. Clearly, a great deal will hinge on the overall buying and macro environment. But naturally, we are still we are expanding more quickly than the industry and more quickly than every one of our competitors. So that's a significant characteristic. Naturally, we'd prefer to grow even faster, but I think considering the start of the year and the political uncertainties that exist at present, I think it's a solid guide.
Right. And Wamsi, from where we sit, I would tack on a few additional items to what you just said. Clearly, the guidance we've issued for this year assumes an IT spending backdrop that stays more or less flat compared with the prior year. So no meaningful or major shift in that backdrop is baked into our thinking. Also, I should remind you that as we look at FY '25 revenue growth, higher Evergreen//One opportunities that converted into traditional sales during the back half of the year gave us some lift, and I mentioned this on last quarter's call. Once you adjust for that tailwind, it comes to roughly a point and a half in FY '25. So, taking that together with what Charlie noted about anticipated Evergreen//One growth in FY '26, when we normalize for those two items, we're looking at somewhat of a step-up in revenue growth compared with the prior year.
Thanks, Wamsi. Let's take the next question.
Great, thank you. So, regarding the CapEx versus OpEx dynamic — a year ago, when the CapEx environment was tighter, you mentioned seeing an acceleration in Evergreen. I just wanted to understand whether you think customers are leaning more toward OpEx — sorry, toward CapEx right now as they refresh a lot of data center equipment, or what dynamics you're observing around customers' willingness to make CapEx versus OpEx purchases? Thanks.
Yes. Meta, we observed considerable change throughout the year. Every quarter came out somewhat different from the one before it and from the quarters in the prior year, which you referenced, and we believe the market remains quite immature when it comes to what is a traditional product — where the corporate data center, the on-prem data center, has historically bought their products — versus where we are among the first to genuinely deliver an as-a-service offering, a SaaS offering on prem. What we find is that this is a more complex sale, and that complexity brings in the finance organization inside a customer, where they need to get their heads around what it means for them, not only from a total cost perspective, but honestly from their own reporting and accounting perspective, which makes it a longer and more challenging sale, whether that's for us or for our channel as well. So we still see quite a lot of misunderstanding and immaturity, if you will, across the market overall. And we think that continues to cause greater difficulty when it comes to achieving consistency, if you will, on a quarter-by-quarter basis. So it's something where we do expect that as the market matures, we'll begin to stabilize more as we move forward, but it does make projecting a little bit difficult.
If I go a bit deeper into Q3, that's when we really had a notable spike, so to speak, in Evergreen//One deals turning into conventional purchases. During Q4 we observed that ease off somewhat, and I believe that's partly what fed into the record sales posted in the quarter too. Then, stepping back to consider how Evergreen//One did overall across the year and what caused that drop versus the prior year, the main factor is really when the bigger contracts landed — we classify those as above $5 million — since they closed during the first half, given that we made up for it in the second half, matching last year's count of large deals in that second half. Meanwhile our velocity business, meaning anything under $5 million, improved substantially; we recorded solid growth annually there. So as I see it, looking across the whole year, everything really traced back to those bigger deals and when they happened to close during the first half.
Thanks, Meta. Let's take the next question.
Oh, wonderful. I appreciate you fitting me in. Regarding gross margins, is it necessary to make some NAND purchases ahead of time to reach the guidance you've given for better product gross margins in the E family? Could you just take us through the basis for your confidence and what supports the belief that product gross margins will rise compared with what you reported in 4Q? Thank you.
Yes. Thanks, Asiya. We genuinely think that — well, to begin with, buying in advance and handling pricing in the market, so to speak, is an area where we truly think we possess a very strong capability. Now, we don't — I realize certain other businesses have made announcements about times they made advance purchases. That isn't something we've ever really done. We're active in the market daily, monthly. We're pretty adept at purchasing ahead when we think it benefits us, or turning to the spot market when we think that works in our favor; so we likely won't begin doing that now. What I'll add is that this is simply something we're constantly in the market for. And certainly, we'll seize, say, opportunities in the market whenever we think they're available.
Thanks, Asiya. Let's take the next question.
Thanks. Apologies, I'd like to return to the topic of pricing. It's fascinating that this was once about NAND pricing. As you mentioned, everybody else faced the same situation, and now you're up against the HDD more directly. That's a different sort of calculation. I'm just wondering how you view things like HDD improvements, and it sounds like perhaps you see 2024 as a one-off kind of year, with NAND going up while HDD didn't. But how do you think about the possibility of that recurring down the road, and what it would mean for how long it takes for NAND to replace HDD? How does that shift the ROI math or anything else — any insight you can offer would be useful. Thanks.
Sure. Thanks, Tim. Tim, you clearly know this market well. Yes, indeed, we compete against HDD systems, correct? So picture a system built on NAND against one built on HDD. I'd add that there are — though we don't want to dive into all those specifics on this call — several distinct tiers of HDD systems, depending on capacity and performance level, with the weakest performance tier clearly being the least expensive option available. With that said, we think that at a system level, HDD systems are essentially priced as low as they could ever go; meaning the HDDs themselves play a smaller role in the overall system cost relative to the quantity of shared equipment required today to operate those environments. So, to put it at length, our pricing and gross margins will hinge far more on NAND prices than on HDD systems, rather than on the HDD price itself. Put differently, HAMR, or higher HDD density, won't matter much at this stage. What will truly decide it is the NAND market. And once we reach both a competitive price and a competitive gross margin, then I believe it's game over.
And Tim, it's Rob here. I just want to build on that point. I believe the initial portion of what you asked was, hey, when you're speaking with the HDD manufacturers, they're presenting their roadmaps, so how does that come into play? Like we've mentioned, we truly haven't observed, and we don't anticipate, the pricing landscape for hard disk drives shifting all that significantly. Nor do we expect the technology side to change very much either. Disks are finding it hard to increase in size, yet they aren't becoming any quicker. In fact, per unit of capacity they're genuinely slowing down, and that's really restricting where they can be used. I think the second portion of what you asked was, hey, so how does this play into the ROI, or really the TCO calculation? And here is where, as Charlie mentioned, when you examine our solutions on a full system basis next to HDD based systems, we're already at a point where you could hand out the disks at no cost, and on a TCO basis, our systems already hold the advantage. So although QLC prices will swing around in the near term, the longer-term direction is obvious; they're going to keep dropping as time goes on. And as that happens, it's simply going to keep enabling us to speed up the replacement of disc.
Thanks, Tim. Let's move on to the following question.
Hello, everyone. I appreciate you fielding my question. You brought up the shifting geopolitical landscape along with tariffs; however, I wanted to ask whether there's any peer-specific factor that might be, that you're currently observing in the market that could put that 11% guidance for next year at risk? Thanks.
Well, one reason this question is hard to answer is that predicting anything particular is extremely challenging. Right now, the political climate truly is highly dynamic and unpredictable; therefore, I'd say nothing specific stands out to us that would deliberately prevent us from reaching the 11%. If anything, in a normalized market — meaning one where we wouldn't anticipate, say, substantial political uncertainty — I'd expect conditions to be getting better over time rather than worsening at this stage.
Right. Once more, regarding our guidance — this is Kevin speaking. For the coming year, when it comes to growth in revenue, we're again assuming an IT spending backdrop that holds steady compared with the prior year. So should that shift upward or downward in a meaningful way, we'd factor it in. Beyond that, something else we pointed out is our expectation that Evergreen//One, which is our service offering, will expand in the next year, and that too is included in the 11% figure we're projecting.
Let's take the next question.
Appreciate you taking my question. Good to be on the line here. Regarding the upcoming GTC conference, is there anything more you can share about why you decided to participate? I mean, it appears somewhat different from what was said previously about going after the training opportunity? And yes, I'm just wondering how you expect to differentiate yourselves in this space? Thank you.
In the past, the AI training market has typically been oriented toward HPC, or high performance computing. Our traditional sales have been into the standard enterprise market, and the values, so to speak, that guide purchasers' choices in those two markets are in fact fairly distinct. We've consistently pointed out that our technology is very superior, thanks to our Purity operating environment and the unique benefits we get from DirectFlash; we've managed to make certain changes to our product that genuinely enable it to tackle many of the performance traits that HPC environments are specifically seeking. And credit goes to our team for managing to accomplish that within our R and D spending budget and for being able to deliver it at GTC. So we're very proud of that. We believe it will create a big splash, and all of you can head to pure.ai right now and begin to be part of the roll up to the to GTC.
Thanks, Ari. Let's take the next question.
Thanks, everyone. Appreciate it, Charlie. Appreciate it, Kevin. My question is about Portworx. In your prepared remarks, you mentioned that 2 Fortune 50 companies moved off VMware with Portworx onto what appears to be OpenShift. Could you give us a bit more detail on what drove that, what those customers were after? You touched on it somewhat in the release, but we'd like more insight into why BoardWorks is succeeding and essentially shifting those environments away from a VMware foundation? Thanks.
Yes. Well, in line with what we've discussed before, a great many customers see Kubernetes and containers as their future solutions; and if Kubernetes is how you intend to manage your application environment going forward and today you're a virtualized solutions customer, as most customers are, with substantial virtualization in place, why would you manage that any differently from how you'll manage your future container environments? So the draw and the appeal, so to speak, of adopting Kubernetes as your foundational application management environment is quite strong; and with what's called KubeVirt, or Kubernetes virtualization, now available, that's an attractive path forward, so to speak, for companies aiming to reach a consolidated integrated application management environment. And I believe that's what has pulled both of those companies in that direction; though to be completely fair, it's — < > remains immature just as every other solution does relative to VMware. Yet over the long term, it's attractive since it's integrated with Kubernetes.
Appreciate that, David. For our final question, we'll go to someone who has come back into the queue. Thanks for doing that.
Hey, I appreciate you letting me ask a follow-up. When it comes to your FY '26 guidance — that full-year outlook — is the one to two exabytes of hyperscaler shipments you referenced previously included in it?
Yes, Howard, it is. For this year's guide, we've factored that in. So there's nothing beyond what we discussed last quarter on that front.
Appreciate that, Howard. As we wrap up here, I believe Charlie would like to offer a few closing remarks.
Absolutely, Paul — thanks. As we wrap up the final quarter of our fiscal year, I'd like to take a moment to acknowledge how much our team has accomplished. Our approach around the platform keeps fueling change across the storage sector, and we're assisting enterprises in tackling their scattered data landscapes right now, while also enabling them to tap into the complete capabilities of AI and create their very own enterprise data clouds. Securing that hyperscale design win marked a significant step forward; in my view, it really underscored the part we play in handling data management at massive scale and the expansion of data, especially when it pushes up power consumption at the hyperscalers. I'd like to express gratitude to our customers, investors, employees, partners, and suppliers for standing behind us all year long, and we're eager to keep moving forward on this path together in FY '26. Thanks.
That wraps up our discussion of Pure Storage's financial results for the fourth quarter and the entire fiscal 2025 year. We appreciate everyone joining us. You're free to hang up at this point.
Read the full transcript via the API
You're viewing the first half of this call. Get the complete Everpure, Inc. transcript - plus 132,000+ transcripts from 7,000+ companies, speaker segments and full-text search - through the EarningsAPI REST API or hosted MCP server.
Get an API key View API docs →For developers and AI pipelines
Programmatic access to Everpure, Inc. earnings transcripts and 132,000+ others is available through the
EarningsAPI REST API and the hosted MCP server.
Quarterly plans from $162 - full transcripts, speaker segments, full-text search,
and the /api/v1/transcripts/recent polling endpoint for ETL pipelines.