This conference call transcript was computer generated and almost certianly contains errors. This transcript is provided for information purposes only.EarningsCall, LLC makes no representation about the accuracy of the aforementioned transcript, and you are cautioned not to place undue reliance on the information provided by the transcript.

Pure Storage, Inc.
2/26/2025
Good day and welcome to the Peer Storage fourth quarter and full year fiscal 2025 financial results conference call. Today's conference is being recorded. All lines will be muted during the presentation portion of the call with an opportunity for questions and answers at the end. If you would like to ask a question, please press star one on your telephone keypad. At this time, I'd like to turn the call over to Paul Zietz, Vice President of Investor Relations. Please go ahead.
Thank you. Good afternoon, everyone, and welcome to Pure's fourth quarter fiscal year 2025 earnings conference call. On the call, we have Charlie Giancarlo, Chief Executive Officer, Kevin Kreisler, Chief Financial Officer, and Rob Lee, Chief Technology Officer. Following Charlie's and Kevin's prepared remarks, we will take questions. Our press release was issued after close of market and is posted on our website where this call is being simultaneously webcast. The slides that accompany this webcast can be downloaded at investor.purestorage.com. On this call today, we will make forward-looking statements which are subject to various risks and uncertainties. These include statements regarding our financial outlook and operations, our strategy, technology, and its advantages, our current and new product offerings, and competitive industry and economic trends. Any forward-looking statements that we make today are based on facts and assumptions as of today, and we undertake no obligation to update them. Our actual results may differ materially from the results forecasted, and reported results should not be considered as an indication of future performance. A discussion of some of the risks and uncertainties relating to our business is contained in our filings with the SEC, and we refer you to these public filings. During this call, all financial metrics and associated growth rates are non-GAAP measures other than revenue, remaining performance obligations, or RPO, and cash and investments. Reconciliations to the most directly comparable gap measures are provided in our earnings press release and slides. This call is being broadcast live on the Peer Storage Investor Relations website and is being recorded for playback purposes. An archive of the webcast will be available on the IR website and is the property of Peer Storage. Our first quarter fiscal 2026 quiet period begins at the close of business Friday, April 18, 2025. With that, I'll turn it over to Charlie. Thank you, Paul.
Good afternoon, everyone, and welcome to our Q4 and fiscal 2025 earnings call. We delivered a solid Q4, exceeding both revenue and earnings guidance, and capping off the year marked by progress as our long-term data platform strategy is changing how enterprises think about data storage and management. In FY25, we delivered on several key strategic initiatives. First, We achieved a design win with a top four hyperscaler. This relationship is proceeding well in the next stages of both testing and deployment plans. We see continued progress on certifying our use across multiple price performance layers. Fiora's direct flash technology will enable this hyperscaler to deploy a consistent architecture across multiple performance tiers of their storage hierarchy. It also allows them to redeploy significant amounts of power and space to new workloads and to increase their data storage performance, reliability, and lifetime. We also launched our 150 terabyte direct flash module, marking a five-fold increase in capacity over standard hard disks and two and a half over the largest SSDs. We are on track to deliver a 300 terabyte DFM late this year. Next, our expansion and growth of the E family during the year signified a major advancement to address the expected tenfold increase in unstructured data driven by AI over the next five years. This offers businesses far better economics compared to disk. along with superior power and density efficiencies that will drive the displacement of disk in data centers. The E family has become a strong competitor to existing hybrid and hard disk environments. The rise of NAND prices in 2024 relative to disk competition did affect gross margins negatively in the year, which we expect to abate in 2025. We also continue to invest and see success in our Evergreen One program, which has attracted strong customer satisfaction and high existing customer growth. Perhaps our most significant and revolutionary technology advancement last year was the GA release of our Fusion version 2. Available as a non-disruptive upgrade to all Pure customers, Fusion v2 transforms traditional enterprise data storage architectures from application data silos to enterprise data clouds. With Fusion, Pure can now provide customers with dataset management services supporting their global data estate. This is an area in the enterprise which is largely managed manually, if at all. Manual management and inadequate traceability of datasets is a leading cause of data sprawl and cyber risk for enterprises worldwide. Fusion opens up significant new opportunity and value for Pure in this space. The best way for me to describe the transformational impact of Fusion is with a familiar analogy. A decade ago, many of us used an external hard drive to add storage to our computers. It provided additional storage and could back up our files. However, when it filled up, we needed to replace it with a larger unit and migrate the data from the old unit to the new. If a colleague also had a hard drive which had spare capacity, it couldn't easily be shared. and file sharing between computers was difficult. Finally, if the external hard drive failed, there was little that one could do to recover. Today, few people use external drives. We've all moved to personal cloud storage provided by many different cloud providers. The benefits go well beyond just storage. For example, personal cloud storage provides Unlimited storage scalability. When one needs more, one only needs to pay for more without having to migrate data. Consistent updates to their infrastructure with non-disruptive software and hardware improvement. In other words, the service doesn't grow old. And guaranteed resiliency. One expects never to lose your data. Global access. One can access their data from anywhere, from any machine, with search features to find content. and easy file sharing with just a few mouse clicks. This model has extended to how colleagues now work in a company. Corporate provisioning of personal data storage on services such as Google Drive or Microsoft OneDrive further allows companies to place their corporate data policy on employee data management, such as controlling file sharing to meet corporate compliance requirements, managing file retention and deletion policies, enabling or disabling individual access when employees join, change roles, or leave the company, and providing administrative global search for all stored files. In stark contrast, the surprising secret about enterprise IT data storage is that it works almost exactly like those old personal external disk drives. Standard storage arrays operate as external storage devices attached to a specific application stack. Two storage arrays in the same data center attached to two application stacks are managed separately and do not share performance or capacity. Many enterprise customers have hundreds of storage arrays globally, each operating as an independent data silo. The amazing advancement of Pure Fusion V2 transforms pure storage arrays from individual storage arrays into an enterprise data cloud. Pure Fusion allows all of an enterprise's pure arrays to be managed as an integrated system. Better yet, it allows the stored data to be managed as a data cloud rather than individual data silos. Fusion allows applications to utilize the total shared capacity and performance of all the arrays, not just the ones that they are directly connected to. Fusion automates compliance policies for secure, consistent data handling. It enhances cybersecurity and eliminates manual errors. It enables automated data set management, enforcing and tracking corporate rules and policies. It also allows stored data to be accessed across the global enterprise, breaking down data silos and enabling a true enterprise data cloud. Early feedback from dozens of customers has been enthusiastically positive. As enterprises face challenges like AI integration, cybersecurity threats, and hybrid cloud optimization, the need to move beyond traditional manual storage management is clear. Operating as a data cloud, Fusion automates data management, enhances security, and simplifies hybrid on-prem and cloud environments. It supports real-time AI access, eliminates the need for excessive copies of data, and fosters API-driven innovation. Pure Fusion enables a fundamental shift from data silos to an enterprise data cloud model for data management. With Fusion, Enterprises have a universally accessible storage solution that can automate and simplify their data management across thousands of arrays as simply as managing one array. This enables enterprise to operate all of their data storage on a workflow basis, just like the cloud. This marks not just an evolution, but a revolution in IT infrastructure, redefining enterprise storage standards and data management. The success of several large enterprise deals this quarter was largely due to this expanded platform strategy. For example, a financial services technology company re-architected its entire storage environment with Pure due to slow response times with legacy architecture. By consolidating fragmented storage onto our unified platform, the company now turns data into a powerful asset while processing hundreds of billions of financial transactions a year. One of the world's largest telcos selected Pure to tackle its significant data center power consumption challenges and reduce energy usage in data storage by an impressive 89%. Our 150 terabyte DFMs were foundational to this success, driving substantial business impact. Our solution enabled this provider to manage petabytes of data, run advanced analytics, and archive data for extended periods with the same core architecture. SmartWorks, our storage solution designed for containerized applications in Kubernetes environments, saw a strong increase during FY25, driven by significant growth in AI analytics and cloud-native platform use cases. Last quarter, two Fortune 50 companies decided to transition their VMware environments to a modern virtualization environment using Portworx and Red Hat OpenShift. One of these companies will significantly accelerate its application deployment speed and time to achieve results while also reducing its cost by 30%. The other, a major manufacturer, is consolidating their containerized and virtualized application platforms in their manufacturing plants onto Kubernetes. Portworx enables them to reduce the cost of managing different infrastructures for VMs and containers and to simplify their operations. As I've indicated in recent quarters, Our activity in AI continues to grow. Machine learning and training environments are increasingly utilizing our high-performance storage, and AI data preparation chains increasingly utilizing Portworx. We are looking forward to showcasing our latest AI advancements at the upcoming NVIDIA GPU Technology Conference next month. We will unveil how FlashBlade will set a new bar for unmatched performance scalability, and ease of deployment for large-scale AI infrastructure deployments. Turning to the macro environment, we expect geopolitical uncertainty to contribute to a dynamic environment at least through this year. In addition to our standard practice of operating with globally distributed and diversified supply and distribution chains, we have also developed contingency plans for a variety of tariff scenarios. Overall, we are pleased with our performance in FY25 and remain highly confident in our differentiated leading data platform strategy and vision. We are excited to help our customers build their own enterprise data clouds across their distributed and hybrid IT environments. This confidence is reinforced by the four key competitive advantages of the Pure platform. Our unified purity operating system, our evergreen technology and business model, our Purity Direct Flash technology, and our cloud operating model, now enhanced with Pure Fusion V2, which allows our customers to build their own enterprise data cloud. With that, I'll pass the mic to Kevin.
Thank you, Charlie. We close the year on a high note, exceeding guidance with double-digit revenue growth, generating $153 million in operating profit, and achieving an operating margin of 17.4%. In Q4, total revenue increased by 11%, and we set a record for TCV sales of Evergreen One. In FY25, we reached a major financial milestone by surpassing $3 billion in total revenue for the first time, with total revenue of $3.2 billion growing 12%. We also delivered our highest annual operating profit of $559 million. Sales of our solutions across our data storage platform contributed to our strong Q4 performance. Record sales in Q4 were achieved by FlashBlade, FlashArray XL, Portworx, our eFamily, and renewals of our Evergreen subscriptions across our install base. Accelerating customers' transition of cost-sensitive workloads to our eFamily and FlashArray C solutions remains a key strategic priority as customers are increasingly recognizing the superior economics over traditional disk solutions. Record sales of our eFamily solutions combined with higher QLC Flash costs and comparatively stable hard disk solution pricing have contributed to temporarily lower product gross margins of 62.9% in Q4. We also closed out the year with record quarterly TCV sales of Evergreen One of $140 million, a 20% increase. FY25 TCV sales for Evergreen One and our other service-based offerings reached 393 million, representing a 3% decline. This was the result of both extended timelines needed to close Evergreen One deals greater than 5 million and a higher conversion of Evergreen One opportunities to a traditional sale that we experienced last quarter. We remain confident in the long-term growth potential of our expanding as-a-service offerings. We expect TCV sales of Evergreen 1 to grow next year, and we will continue to provide quarterly updates, though specific guidance will not be provided. Q4 subscription services revenue of $385 million increased 17%, and subscription services annual recurring revenue, or ARR, grew 21% to $1.7 billion. Total remaining obligations or RPO exiting Q4 encompassing both subscription services and non-cancellable product orders grew 14% to 2.6 billion. Excluding non-cancellable product orders, RPO related solely to our subscription services increased by 15%. Overall RPO growth was impacted by the decline in TCV sales for our storage as a service offerings during FY25. Our subscription services net dollar retention, or NDR, at the end of the year was 117% and continues to be aligned with our long-term target of 115%. In Q4, U.S. revenue of $619 million was the primary driver of growth, while international revenue reached $261 million, down 3% year over year. For FY25, U.S. revenue grew by 12%, and international revenue increased by 13%. Additionally, we acquired 334 new customers, generally consistent with both Q4 FY24 and the previous quarter. We continue to serve 62% of the Fortune 500. In Q4, total gross margin was 69.2%. supported by a robust subscription services margin of 77.2%, while product gross margin, influenced by record sales of our E family, stood at 62.9%. For the year, total gross margin was 71.8% compared to 73.2% in FY24. Looking ahead for the full year of FY26, we expect product gross margins to settle in the mid-60s, which is in the range of our long-term expectation for product gross margins of 65% to 70%. This outlook is driven by expected strong demand for our e-family and flash array C solutions, alongside our expectation that QLC flash pricing will moderate, an important factor as we compete against disk-based solutions. Our headcount increased slightly to nearly 6,000 employees at the end of the year. Our balance sheet remains robust with 1.5 billion in cash and investments at the end of the year. Cash flow from operations in Q4 was 208 million and 753 million for the year. Capital expenditures during the year were 227 million, representing approximately 7.2% of revenue for FY25. Our capital investments during the year supported data center expansion to support expanded testing of new products and solutions, including a recent hyperscale solution design win. Accelerating density of our direct flash modules and software development of our Fusion version 2 solution. Free cash flow for Q4 was 152 million and 526 million for the year. Free cash flow margin for the year was 16.6%. In Q4, we repurchased 3.1 million shares, returning approximately 192 million to shareholders. For FY25, share repurchases amounted to 374 million, or 6.7 million shares. Additionally, we paid 65 million in withholding taxes on employee awards in Q4, offsetting dilution by about 1 million shares, and 209 million for FY25, offsetting about 3.5 million shares. We currently have 21 million remaining under our existing repurchase authorization and are announcing a new share repurchase program for an additional 250 million. Turning to our FY26 guidance, we again expect double digit revenue growth, increasing by 11% to just over 3.5 billion. Our annual revenue outlook considers an IT spending environment similar to FY25. and a return to growth for Evergreen One and other service offerings. For Q1, we expect revenue of $770 million, representing an 11% year-over-year increase. Moving to operating profit, we remain committed to our bold strategic initiatives, positioning ourselves for strong, sustainable long-term growth. This includes leveraging our recent hyperscale design win through incremental investments to scale operations for large production and deployments starting in FY27, as well as accelerating density in our direct flash modules. We expect an operating margin in FY26 of approximately 17% consistent with my remarks last quarter, translating to an operating profit of around 595 million. For Q1, we are guiding operating profit of 80 million and operating margin of 10.4%. In closing, we are pleased to have returned to double-digit revenue growth and surpassing 3 billion in revenue for the first time. This pivotal year has been marked by remarkable innovation and execution, setting the stage for sustainable long-term growth. With that, I will turn it back to Paul for Q&A.
Thanks, Kevin. Before we begin the Q&A session, I'll ask you to please limit yourselves to one question consisting of one part so we can get to as many people as possible. If you have additional questions, we kindly ask that you please rejoin the queue and we'll be happy to take those additional questions as time allows. Operator, let's get started.
Thank you. If you would like to ask a question, please press star followed by one on your telephone keypad. If for any reason you would like to remove that question, please press star followed by one. Again, to ask a question, press star one. As a reminder, if you're using a speakerphone, please remember to pick up your handset before asking your question. We'll pause here briefly as questions are registered. Our first question comes from Mike Sikos from Needham. Please go ahead, your line is open.
Great. Thanks for taking the questions here, guys. I wanted to come back to some of the prepared commentary on Fusion. And really, I know it's early with the launch of Fusion 2.0, but do we have signs yet of customers that have adopted Fusion showing either better expansions or stickiness, or if it's still too early, can you at least provide commentary on the go-to-market potentially and where we are in driving the adoption and awareness of what Fusion enables? Thank you.
Thanks, Mike, uh, Charlie here. Uh, thanks for the question. So, you know what we've had now in order for a customer to experience fusion, they have to, uh, download the latest update to the, uh, software, uh, operating system in, in flash array. Uh, we've had dozens of companies do that, uh, and specifically turn on fusion and start operating with it. And with less than a quarter of that being in the, in the market, uh, you know, as I said, we've had dozens of customers do that. That's a very fast uptake. And as I mentioned in my prepared remarks, of course, we're getting feedback. And the feedback has been not just positive but enthusiastic. They find that the ability to manage the systems that they put under Fusion as a fleet rather than as individual arrays to make life just a lot easier. They are now experimenting with something we call presets. What presets allow them to do is set policies for how the data is handled across the arrays. such that for any specific storage class, all of the data in that storage class will be handled the same way. So as opposed to having to manage different application use cases manually in terms of policy, now the policy is carried out under software control. So the feedback's been very positive. Obviously this creates a network effect. It creates a network effect for our products, If you have all or even some significant portion of your pure storage arrays under fusion, the likelihood of you adding another array that's not part of that I think is going to be pretty low. But to your point, it is a bit early at this point. We're still in the early rollout phases. I would expect that type of network effect to occur after several quarters of use.
Thank you, Mike. Next question, please.
Our next question comes from Aaron Rakers from Wells Fargo. Please go ahead. Your line is open.
Yeah, thanks for taking the question. I wanted to ask about the hyperscale opportunity. And while I can appreciate it's early to really unpack the announcement from last quarter, I'm curious, you know, any progression of how maybe the model around this opportunity might be evolving and then on that same kind of narrative, you know, how would you characterize the discussions or the degree of discussions you might be having with other hyperscale opportunities? Thank you.
Yeah, as you might imagine, we are early, but the conversation continues to evolve and to expand, frankly, in terms of the use cases for different types of technologies. of data storage tiers inside that hyperscaler. So we've been evolving and continuing in terms of the depth of the testing and the co-development in particular, but also discussions around future states and where that storage will go. Relative to other vendors, we continue to expand our conversations. I would say we indicated last time we expected them to accelerate. accelerated through this quarter. Nothing to announce just yet, but certainly, you know, we're seeing increased interest. Rob, do you want to add anything?
No, I think that summarizes it well. You know, the engagement is progressing as we expect and, you know, as Charlie mentioned, moving forward through the more advanced phases of completing their kind of implementation plans and detailed testing and The typical engineering process we would expect to see leading to more detailed field testing on the way to ramping the significant scale. And yeah, I mean, the discussions and engagements we're having with other hyperscalers are definitely moving forward with a faster pace.
Thank you, Aaron. Next question, please.
Our next question comes from Amit Daryanani from Evercore ISI. Please go ahead. Your line is open.
Thanks, Doug. Good afternoon, everyone. I guess maybe just to follow up on the hypothetical question from Aaron, can you just talk about, you know, what investments are needed by Pure at this point to get this from a design win and have it progress to a production environment? I would love to understand kind of what investments do you need to make to get there? And then just any clarity on what steps are needed to actually move this from design today to, I think, starting production in fiscal 27. That would be helpful to understand. Thank you.
Yes. Um, so this is a code, uh, ongoing code development with that, uh, with that hyperscaler, meaning that they are developing, you know, new software, new architectures, uh, new infrastructure architectures for, uh, as we've discussed, it's their next generation data center design. And as they design that, uh, that software and those hardware architectures, uh, we have to work alongside them to make sure that our software works closely with theirs. of course we're constantly working together to try to improve both performance and price performance uh in that environment as we mentioned we're working now on multiple uh different performance tiers each one of them requires its own set of tuning and of course we have to also qualify um several different scales of our direct flash modules along with several different flash partners, if you will, different flash manufacturers for those direct flash modules. All of that requires quite a bit of work and testing on our side as well.
And this is Kevin. I'd probably summarize it into four key areas just to summarize what Charlie's points are. integrating the Purity software and our direct flash technology with the hyperscaler hardware specifications, accelerating our density roadmap with our direct flash modules, qualifying additional NAND suppliers that Charlie alluded to, expanding our supply chain capabilities and scale, and really, as well, there's investment in pursuing other opportunities, hyperscaler opportunities as well.
Thank you, Amit. Next question, please.
Our next question comes from Howard Ma from Guggenheim. Please go ahead. Your line is open.
Thank you. For Charlie and or Kevin, can you help us understand how much of the product gross margin decline is, how much is it due to incremental sales of the E and the C families versus, you know, say higher KLC costs or perhaps more aggressive pricing in other areas of your portfolio?
Thanks.
Yeah. So yeah, our gross margin performance on other areas of our portfolio has stayed relatively constant. So no real change there. So it's entirely attributable to the E family. Now there are two effects of that. One is that the E family now is a higher percentage of our overall mix. But the other effect is that the E family Um, because it competes with disc that, uh, you know, where the rise in NAND prices in the media prices did not affect the disc market. So in that case, you know, whereas when we compete with other flash pricing, our margins have been, have been very stable, but when we compete against disc solutions where NAND prices have increased, it affects that family more as you might imagine. So we view that as, uh, as somewhat temporary as NAND prices moderate this coming year. where the margins for the E family will come back 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 start to approach our normalized margins. But as we indicated many quarters ago, we were going to be aggressive with the E family, given that we're really the only player in the market right now that can compete with DISC, and we want to take advantage of that. Uh, so we are being aggressive there. The, the more recent last year's uptick in NAND, uh, pricing, uh, you know, affected us a bit more than, uh, you know, one would have hoped.
Yeah. And so with that, this is Kevin, uh, Howard, uh, you know, just to, to follow on to Charlie's points, as we think about next year, we expect fully that the e-family momentum will continue. Uh, but with the moderation of QLC pricing. We do think we'll get back into a product gross margin range of around 65%, which, again, is aligned with our long-term product gross margin profile.
Thank you, Howard. Next question, please.
Our next question comes from Pendulum Bora from JP Morgan. Please go ahead. Your line is open.
Hey, guys. This is for Pendulum. Thanks for taking our questions. Can you just maybe provide a little color about some of the geographical theaters given the outperformance in the U.S. relative to international this quarter? Thank you.
Yeah, this is Kevin. Look, I think one is we were actually quite pleased to see the performance in the U.S. and the growth driven by the U.S., not unusual to see variability in growth by geographies on a quarterly basis. I think on balance, you know, the growth was actually relatively comparable for the year, both for international markets as a whole, as well as the U.S. And look, our international revenue growth was quite strong through the first three quarters of the year, came down obviously in Q4, but nothing that I would call out specifically. And We do expect our international markets and revenue to grow at a faster pace longer term than the U.S., given the sizing of the markets and opportunity.
Thank you. Next question, please.
Our next question comes from Chris Sankar from TD Cowen. Please go ahead. Your line is open.
Hi, thanks for taking the question. Question for Charlie or Kevin. You know, on RPO, it looks like the RPO growth rate has declined from about 30% a year ago to 14% in the last quarter. I'm kind of curious. Is this just a law of large numbers or is there something else going on the RPO and what does it mean for your core enterprise business? Thank you.
Yeah, this is Kevin. I'll start out and then have Charlie add. I mean, this is just really a derivative of what we saw in our evergreen one TCV sales this year. So obviously with a small drop in revenue for the year, that's impacting our RPO for the year as well at 15% when you exclude non-cancellable product orders. But, you know, look, we were quite pleased to see in Q4 a record Evergreen One performance of 20%. We've certainly considered growth in our revenue guide for this year of our Evergreen One offering. And we're very pleased and excited with that offering and the value realization that our customers are seeing with Evergreen One as a whole.
I got covers, Kevin.
Thank you. Okay. Thank you, Chris. Next question, please.
Our next question comes from Jason Ader from William Blair. Please go ahead. Your line is open.
Yeah, thanks. Good afternoon, guys. I wanted to just follow up on Raker's question earlier. I don't think, Charlie, you answered the question on just the model and how gross margin and operating margin might look for the hyperscaler business. And do you think you'll be able to provide more specific guidance on on revenue as we move through the year?
The answer on that is yes. As we understand better the exact rollout of the hyperscalers deployment plans, and we assume that we should be able to understand that in the three to six months prior to the actual rollout, we should be able to give an update on exactly where that stands. I'll say that right now it's fairly dynamic. So until we get a little bit on firmer footing, dynamic just in the sense that we're still in the planning phase and a lot of considerations in terms of the exact timing. So as soon as we get into a place where we feel more confident about the exact timing and the rollout, we'll be able to give more updates.
And Jason, the other thing too that I would add from a financial perspective is really nothing new from what we just discussed last quarter in detail with that design win. You know, from a gross margin profile, again, hardware would not be reflected as how we're planning that. Obviously, a longer term from an operating margin, we expect expansion over time with this opportunity. But again, these factors are very consistent with what we described last quarter in detail.
Thank you, Jason. Next question, please.
Our next question comes from Simon Leopold from Raymond James. Please go ahead. Your line is open.
Thanks. I was hoping you could update us on your thinking of how enterprise adoption of AI initiatives could affect your business. I know, Charlie, you've highlighted this uplift phase If you could talk about how you think that could affect your business and when. Thank you.
Yeah. Well, AI, as we've talked in the past, is a multifaceted subject. One is there's the large-scale machine learning environment, which is really just the top, you know, the peak of the pyramid, if you will, of enterprises that will be doing their own machine learning environment. We did hint, I did hint, that we've got a... a nice new announcement coming that we'll be demonstrating at GTC. That really does relate to the largest scale of the AI opportunities. Second is the inference opportunity. We've had a number of conversations. We continue to have conversations with both customers as well as with NVIDIA on the inference opportunity. I still see that as a little ways out, but we're starting to build some momentum early on in terms of early conversations with customers who plan on deploying in that area. I'll state once again that I think our biggest opportunity is in reorganizing the way that customers manage their production data from data silos into an enterprise data cloud. We see that as a major, we see AI as a major prod, if you will, towards customers rethinking the architecture of their data storage into something that is much friendlier, if you will, to AI getting access to data for real-time analysis, especially in inference and RAG-type environments. So we believe that Fusion couldn't be coming too soon, that it's a major opportunity for us, not only in the in the management space, in the compliance space to allow customers to have a better non-manual automated method for managing their data, but frankly, in the AI space, making all of that data available for AI analysis.
Yeah, and Simon, this is Rob, just to add on to that. I think if we step back, the AI segment and the opportunity, I think it continues to be playing out much as we had expected, as Charlie alluded to. You know, you've got the high performance training infrastructure environments that have, you know, in the last year or so, gotten a lot of the attention. What we're starting to see a bit more driving our engagements are as enterprises move more towards the inference or RAG deployment, you know, type phases or start considering how they're going to connect their broader data environments and, you know, data sets and silos into those AI workflows. coming to Pure and Pure Fusion to go help them do that. And so if we step back and we look at where we sit in the AI segment today, a lot of that has been historically led by FlashBlade. But as the inference opportunity starts to play out and certainly as Fusion comes into play, we're also now seeing more and more customers that are deploying our Portworx and FlashArray solutions, really our entire portfolio and platform of solutions to help them drive their AI initiatives as you know, they broaden out from purely just looking at training to the entire gamut of those environments. Thank you, Simon.
Next question, please.
Our next question comes from Jim Fish from Piper Sandler. Please go ahead. Your line is open.
Hey, guys. This is Quinton Umber, Jim Fish. Thanks for taking our question. Maybe two kind of interrelated ones on the margin side for you, Kevin. How, if at all, have you embedded tariffs into this guide And then I think I heard you say that investments for this year are to support other hyperscaler opportunities. So should we assume that any potential incremental announcement this year wouldn't require the incremental spend so our kind of 17% margins that we're establishing are roughly safe? Thank you.
Yeah, that sounds great. Which question would you like to pick? I'll hit the hyperscaler. Yeah, look, I mean, we're building out the ecosystem on the hyperscaler environment. That includes enhancing for scale our supply chain overall to absorb not only increased volume from the one design win, but also new wins as well. But that wouldn't preclude us from requiring investment to grow other opportunities. I mean, each opportunity we're looking at has, um, some customization attached to it in terms of the integration of our purity software, uh, with hardware. Um, but, uh, obviously a huge, uh, long-term opportunity as we see it going forward.
You want to just comment on it's a little bit on the tariff side of things. Um, you know, as we, first of all, I think we've proved ourselves during the COVID period with the, you know, remarkable supply chain disruptions. of being very flexible and capable, uh, and having a solid and reliable, uh, and diverse supply chain. Uh, as you may recall, or for those that haven't followed us for as long, uh, we, uh, at most we went to a six week lead time, which was, you know, leadership in, uh, not just in our industry, but across the tech landscape. Uh, we've further diversified, uh, since then, uh, our, um, our supply chain and our distribution chain. Now, I can't predict nor do I think anybody can predict right now exactly what tariffs are going to occur and when and how long they'll last. We are developing contingency plans for many different scenarios. We will be monitoring tariffs and at the appropriate time, we'll take the right action and we'll report back when we do.
Next question, please.
Our next question comes from Wamsi Mohan from Bank of America. Please go ahead. Your line is open.
Yes, thank you so much. Carly and Kevin, you just noted that E-Series is becoming a larger piece of the mix and you intend to be more aggressive with the E-family. So as you penetrate this opportunity, why are you not projecting a more meaningful reacceleration in revenues? I mean, your exit rate this year is is 11% growth, you're forecasting 11% growth, all things equal from a macro standpoint, but you're really aggressively attacking this opportunity. So is there some other offset? Like, how should we be thinking about sort of the lack of reacceleration in revenue terms? Thank you.
Yeah, if you consider the fact that we also are expecting a return to growth in the Evergreen One area, I think it's a, we think it's a reasonable um, forecast or guide for the year. You know, obviously it'll depend a lot on, on the, uh, the overall buying and, and macro environment. Uh, but you know, of course we, we are, uh, still we're growing faster than the industry and growing faster than all of our competitors. Uh, so, you know, that's a important characteristic. Uh, obviously we'd like to, to grow even faster, but I think, you know, given the beginning of the year and the, and the, uh, political uncertainties that exists right now, I think it's a good guide.
Yeah, and Wamsi, I'd add a couple other points to that from our view is obviously our guide for this year is contemplating an IT spend environment that's relatively consistent year over year. So we're not considering any significant or substantial change to that environment. And I do want to remind you, you know, when we think about the FY25 revenue growth, we did see some benefit. as a result of higher evergreen one opportunities converting to traditional sales in the back half of the year. And I talked about that last quarter. You know, when you normalize for that tailwind, that's about a point and a half in FY25. So we are, you know, and then combined with Charlie's point of expected growth of evergreen one in FY26, we are seeing a bit of an uptick in revenue growth year over year when normalizing for those two points.
Thank you, Wamsi. Next question, please.
Our next question comes from Mita Marshall from Morgan Stanley. Please go ahead. Your line is open.
Great. Thanks. Just kind of on kind of CapEx versus OpEx dynamic, you know, last year when we were in a more constrained CapEx environment, you were talking about kind of seeing an acceleration in Evergreen. I just wanted to get a sense of, you know, do you feel like customers are kind of opting more for capex in this period as they kind of refresh a lot of data center equipment or just kind of some of the dynamics that you're seeing just in terms of customers' appetite for capex versus opex purchases? Thanks.
Yeah, you know, Nita, we saw a lot of change during the year. You know, each quarter turned out to be a little different than the quarter before and different than the quarters last year, which you pointed out, to which we believe that the market is still quite immature in the area of what is a traditional product where the corporate data center, the on-prem data center, has typically purchased their products to where we are one of the first to really provide an as-a-service offering, a SaaS offering on-prem. And what we find is that it's a more complex sale and the complexity involves the finance organization within a customer where they have to get their heads around what it means to them, not just from a total cost standpoint, but frankly, from their own reporting and accounting standpoint, which makes it a longer and more challenging sale, whether that's for us or for our for our channel as well. So we see still quite a lot of misunderstanding and immaturity, if you will, of the market overall. And we think that's still causing, you know, a greater difficulty in terms of getting consistency, if you will, on a quarter by quarter basis. So it's something that we do expect that as the market matures, we'll start to stabilize more as we go forward. But it does make it a little bit difficult to project.
If I break that down a little bit more to Q3, when we saw really a significant peak, if you will, in terms of Evergreen One opportunities converting to a traditional sale, we saw that moderate in Q4. And I think that's in part contributing to the record sales that we saw in the quarter as well. And then if I take a step back on our Evergreen One performance as a whole in the year and look at What was the driver of the year over year decline? It really comes down to the timing of the larger deals, which we defined as over 5 million in the first half closing, because we picked that up in the second half, closed the same amount of large deals in the second half as we did last year. And our velocity business, which is below 5 million, really picked up nicely. We saw a nice annual growth over that. So I really do think that as we look through the year, it was really coming down to the larger deals and the timing of closing of those larger deals in the first half.
Thank you, Mita. Next question, please.
Our next question comes from Asia Merchant from Citigroup. Please go ahead. Your line is open. Oh, great.
Thanks for squeezing me in here on gross margins. do you need to do some NAND pre-buys here in order to kind of get your guidance for improving product gross margins for the e-family just if you could walk us through you know the confidence and what's underpinning the confidence that product gross margins improve here from what you delivered in 4Q thank you
Yeah, thank you, Asiya. We actually believe that, well, first of all, pre-buys and managing, if you will, pricing in the market is something we really believe that we have a very high skill at. Now, I understand that some other companies have announced when they've done pre-buys. We've never really done that. We're in the market every day, every month. We're quite skilled at buying ahead when we believe it's advantageous to us. or using the spot market when we believe that's advantageous. And we probably won't start here. What I will say is that we're in the market all the time for such things. And yes, we'll take advantage of, let's say, market opportunities when we believe we have them. Thank you, Asya. Next question, please.
Our next question comes from Tim Long from Barclays. Please go ahead. Your line is open.
Thank you. Sorry. I just wanted to circle back on the, you know, kind of pricing. It's interesting. It used to be NAND pricing. As you said, everyone else had the same thing. Now that you're competing more with HDDs, it's a different kind of calculus. Just curious how you're thinking about like advancements in HDD. And it sounds like maybe you think this in 2024 was kind of a one-off year where NAND increased and HDDs didn't. how are you thinking about the potential for this happening again in the future and what that would mean for the timeline for HDD replacement with NAND? How does that change the ROI calculation or anything, any call you can provide on that would be helpful. Thank you.
Sure. Thank you, Tim. Tim, so you're very savvy, you know, in this market. You know, indeed, we are competing with HDD systems, right? So think of it as a, NAND-based system versus an HDD-based system. I'd also say that there are, you know, although we don't want to get into, you know, all that detail on this call, there are multiple different layers of HDD systems, you know, depending on size and level of performance, you know, with the lowest level of performance obviously being the cheapest that are out there. That being said, You know, we believe that we are basically, that the HDD systems are basically at a system level about as low a price as they can possibly get. That is to say that the HDDs themselves are not as much of a factor in the price of the total system compared to the amount of common equipment now that it takes to be able to run those environments. So that's a long way of saying that it really are pricing and gross margins are going to depend a lot more on the NAND prices versus the HDD systems, then it's going to depend on the HDD price. In other words, a greater density of HDDs is not going to make that much of a difference at this point. It's really going to be determined by the NAND market. And once we get to a competitive both price as well as competitive gross margin, then I think it's game over.
Yeah, and Tim, this is Rob. Just to add on to that, I think the first part of your question was, hey, as you talk to the HDD vendors, they're laying out roadmaps. How does that factor in? As we've said, we really haven't seen nor do we foresee the pricing environment for hard disk drives to move very much. We don't feel like the technology environment is going to move that much either. Disks are struggling to get larger. But they're not getting any faster. In fact, they are actually getting slower per unit capacity, and that's really limiting their applicability. You know, I think the other part of your question was, hey, so how does this factor into the ROI or really the TCO equation? You know, and this is where, as Charlie alluded to, if you look at our solutions on a full system basis as compared to HDD-based systems, you know, we're already, you know, you could give the disks away for free, and on a TCO basis, our systems are already advantaged. And so, you know, while QLC prices will fluctuate in the short term, the long-term trajectory is clear. They're going to continue to decline over time, and, you know, as that does, that's just going to continue to allow us to accelerate the replacement of disks.
Thank you, Tim. Next question, please.
Our next question comes from Max Michaelis from Lake Street Capital Markets. Please go ahead. Your line is open.
Hey guys thanks for taking my question yeah you know you mentioned a dynamic geopolitical environment, as well as tariffs, but I was wondering if there's anything pure specific you know that may be that you're seeing on the market right now that may risk that 11% guide for next year thanks.
Well, part of part of the difficulty of answering the question is it's it's very difficult to predict anything specific. the political environment is, in fact, very dynamic and unpredictable at the moment. So I would say we don't see anything specific that would artificially block us from getting to the 11%. If anything, I would say that in a normalized market, that is a market where we didn't expect, let's say, a lot of political uncertainty, over time I'd expect the market to be improving rather than declining at this point.
Yeah, and again, our guide, this is Kevin, our guide for next year on revenue growth, again, considered a consistent IT spending environment year over year. So if that changes to the positive or negative in a substantial way, that would be a consideration. And the other thing we highlighted as well is that we are expecting Evergreen One, our service offering, to grow next year, and that's built into our expectations of 11% as well.
Next question, please.
Our next question comes from Ari Terjanian from Cleveland Research. Please go ahead. Your line is open.
Hey, all. Thanks for taking the question. Nice to speak here. Just on the coming GTC conference, like, is there any more color on the decision to, you know, enter here? I mean, it seems a little bit different versus prior commentary in terms of pursuing you know, the training opportunity and yeah, just curious, you know, how you hope to stand out in this market. Thank you.
Well, the AI training market has traditionally been HPC or high performance computing oriented in the past. Yeah, we have traditionally sold into the standard enterprise market and the values, if you will, by which purchasers make decisions, in those two markets are actually quite different. We have always indicated that we have a very superior technology by virtue of our Purity operating environment and the unique advantages that we have with DirectFlash. We've been able to make some modifications to our product that really allows it to address a lot of the performance characteristics that HPC environments are specifically looking for. And, you know, kudos to our team to have been able to do that, you know, inside of our R&D spending budget and to be able to deliver that at GTC. So we're quite proud of it. We think it's going to make a big splash. And you can all go visit pure.ai right now and start to be part of the roll-up to GTC.
Thank you, Ari. Next question, please.
Our next question comes from David Vogt from UBS. Please go ahead. Your line is open.
Great, thanks guys. Thanks, Charlie. Thanks, Kevin. I want to ask a question on Portworx. You referenced in your prepared remarks that you had two Fortune 50 companies migrate from VMware using Portworx to, it looks like, OpenShift. Can you maybe drive a little detail on what was the reasoning behind it? What were the customers looking for? You mentioned a little bit in the release, but we'd love to get some more color on how Portworx is successful and and sort of migrating those environments over from VMware backdrop effectively. Thanks.
Yeah. Well, consistent with our conversations in the past, many, many customers are looking towards Kubernetes and containers as their solutions in the future. And if you're looking to Kubernetes to manage your application environment in the future, and you are a customer today of virtualized solutions, which most customers are, they have a lot of virtualization. Why would you want to manage it any differently than the way you're managing your future container environments, so the attraction and the allure, if you will, of moving to kubernetes as your base application management environment is very compelling. And now that cube what's so called cube vert or kubernetes virtualization now that that exists it's a it's a compelling future. movement, if you will, for companies looking to get to a consolidated, integrated application management environment. And I think that's what's drawn both of these companies towards that direction. Now, to be very fair, Kuber is still immature, just like every other solution compared to VMware, but on a long-term basis, it's compelling because it is integrated with Kubernetes.
Thank you, David. We'll take our last question from a person who re-entered the queue. Thank you for that.
Our last question will come from Howard Maugh from Guggenheim. Please go ahead. Your line is open.
Thanks for taking the follow-up. Does your FY26 guidance, the full-year guidance, bake in the one to two exabytes of hyperscalar shipments that you mentioned last time?
It does, Howard. We've contemplated that in terms of our guide for this year. So nothing additional to that, what we talked about last quarter.
Thank you, Howard. Before we conclude, I think Charlie has some final comments to make.
Yeah. Thank you, Paul. As we close out our last fiscal year, I do want to reflect on the great progress that our team has made. The platform strategy is continuing to drive transformation in this storage industry. And we're helping enterprises address now their fragmented data environments and unlock the full potential of artificial intelligence and to build their own enterprise data clouds. The hyperscale design win has been a key milestone. It's highlighted, I think, our role in solving large-scale data management and growth of data when it drives energy usage in the hyperscalers. I want to thank our customers, our investors, our employees and partners, and our suppliers for supporting us throughout the year, and we look forward to continuing the journey together in FY26. Thank you.
That concludes the Peer Storage fourth quarter and full year fiscal 2025 financial results. Thank you for your participation. You may now disconnect your line.