Thank you, everybody, for joining. For NetApp, first, let me read the safe harbor here that they asked me to read. Today's discussion may include forward-looking statements regarding NetApp's future performance, which are subject to risk and uncertainty. Actual results may differ materially from the statements made today for a variety of reasons described in NetApp's most recent 10-K and 10-Q filed with the SEC and available on their website at netapp.com. NetApp disclaims any obligation to update information in any forward-looking statements for any reason. So again, thank you for joining us. Tim Long here, IT hardware, communications equipment analyst at Barclays. Happy to have Jeff Baxter with us, VP of Product Marketing at NetApp today. Jeff, in this role, is responsible for core product and solutions marketing at NetApp. He's held a number of positions within the company. So, Jeff, thank you for coming.
Maybe we'll start out with some on the higher-level side here. Maybe if you could just talk about, you know, kind of strategic priorities for your business as you head into 2025. What are the, you know, two or three key focus areas for NetApp?
I think certainly, as you'd expect, NetApp is clearly focused on the all-flash market and the growth in the all-flash market. Especially, we've seen significant growth in what we call the capacity flash market, QLC-based sort of media. So we know that's a strategic priority. Within that, we've got a real focus on block storage and a good and growing block storage, specifically with our NetApp ASA product line, which allows us to have greater penetration into the block storage market than perhaps historically we have. Obviously, on the cloud storage side, we think that's a real differentiator for NetApp. We have a healthy and growing on-premise enterprise storage business. And we also have a first-party native presence in the top three hyperscalers in the world. And that mix of it is really what attracts a lot of our customers.
And then, obviously, and last but obviously not least, you'd be surprised, we're focused on AI, but the, you know, AI is obviously a key strategic priority for us in driving both the usage of AI within our products as well as driving AI into our customers for generative AI use cases and more traditional analytics.
Maybe we'll just pivot to AI to start. I think you guys have mentioned 100 AI and data lake modernization wins. Can you talk a little bit more about, you know, a little more detail on who's, you know, what type of customers, what type of applications, any insight you can give us on the early success in AI?
And as you said, we've had a consistent sort of drumbeat of increasing number of wins each quarter. We've talked about some customers on our earnings calls. I think the important thing for us is it's really across verticals, across industries. It's not necessarily completely concentrated in any one area. We've had a sort of broad base of AI wins. The majority of them are. There's certainly some that are in the training sort of space. So training large language models or increasingly training small language models or fine-tuning existing LLMs. But we've seen a lot of growth in inferencing and retrieval-augmented generation. NetApp, having been around in this industry for 30 years, has a tremendous amount of unstructured data in our customers' data estate on NetApp.
And so they're really unleashing that to do things like retrieval- augmented generation using a lot of these different open-source tools and using tools from NVIDIA and others to drive and get access and be able to drive insights from that sort of data. And so we're just picking up those wins quarter by quarter with customers who are either net new NetApp customers or existing NetApp customers who are expanding to AI workloads and taking advantage of their unstructured data.
What has NetApp had to do from a product portfolio or marketing or go-to-market standpoint, to better address GenAI, specifically?
So I'll, I'll take that in two parts. I think from the go-to-market perspective, we've broadly enabled our sales force to be able to pursue these AI opportunities. We, it's sort of a natural extension of the conversations we are already having with customers about what we call intelligent data infrastructure. GenAI and AI in general is kind of a natural extension. We have put together an AI specialist team on our go-to-market that really exists to take the second or third call and really go into depth with customers and be the true data experts.
We talk about a lot of our customers, once they get past the initial look at AI and they look at it as a, you know, I've got to get GPUs, I've got to get, you know, that sort of environment. They realize that the toughest problem with AI problems are data problems, right? And so we have these AI sales specialists and technical specialists who are able to come in and help customers unlock the value of their data. So there's some investment on the go-to-market side there. There's clearly investment from the marketing side in terms of continuing to build our presence and build our awareness of our AI solutions.
Then the product portfolio side, so far we found our existing portfolio, our unified data storage on both on-prem and cloud, has been remarkably resilient in terms of being able to serve AI workloads and GenAI workloads. Specifically, a lot of AI workloads start in the cloud. They may continue in the cloud. A lot of them, once they move past the experimentation stage, large enterprises may move on-prem. Being able to have a presence in all the major clouds and on-prem has been a huge value to our portfolio. We talked about, at our user conference, which was called NetApp Insight back in September, sort of a vision for how we'll continue to evolve our infrastructure to be even more performant and even more available for generative AI workloads. But we've already had tremendous success with the portfolio today.
We don't think we have to fundamentally alter it to take advantage of the AI opportunity.
So what are the changes that, you know, will be evolving as AI gets deeper? Is it, you know, software changes to ONTAP? Is there new product, hardware capabilities that NetApp needs? What needs to change to keep pace with the overall AI market?
So there's been a couple different things. I think we've certainly embedded AI directly into ONTAP most recently for anti-ransomware technologies, right? So that's one way we're using AI in it. In terms of addressing the AI market, the vision we laid out, and again, it's forward vision and subject to change, but the vision we kind of publicly laid out was around evolving ONTAP into really three different layers. One, introducing a disaggregated storage architecture. We know that our existing NetApp AFF all-flash products have won in all sorts of generative AI and large language model training environments.
In fact, we announced with NVIDIA that our AFF product lines are in the middle of being certified with their SuperPOD architecture, which basically says it's, it's among the fastest storage in the world, able to support the fastest AI workloads in the world. We don't want to stop there. We know that the performance demands, you know, with, with Blackwell and with everything else coming, the performance demands are just going to continue to increase. We're introducing a disaggregated architecture into ONTAP. The important thing there is we've said that it's still ONTAP. It's still compatible for our customers. It's still the same API. It's still the same data services. We just allow it to scale out and take better advantage of all-flash performance to generate even higher throughput out of similar hardware. That's step one.
The second step is really building an AI data engine directly into ONTAP. So we think there's an opportunity to really create a new way of interacting with data. So traditionally, our unified storage systems do block storage. They do file storage. They do object storage. We're starting to embed a global metadata catalog directly into our storage along with embedded vector databases. This is our vision to do this so that you start to have almost vector database as a service coming out of the storage layer directly attached to high-speed storage over RDMA so that we can get incredibly efficient vector databases that are always up to date with your enterprise data that you can plug into the GenAI framework of your choice.
The final piece to that is just providing data exploration tools that can leverage that metadata catalog to provide customers greater insight and natural language queries to their data estate on their own.
Great. That's helpful. Maybe we'll touch on public cloud a little bit now. You know, last quarter was very strong. I think the guidance called for double-digit growth in the H2 . You know, this is a business that a few years ago had some up and downs, and now it's been consistently strong. So maybe just talk about kind of what's driving this success, this, you know, this time around. I know there's been some, you know, rejiggering of offerings, but maybe if you can just give us a sense as to what's driving the strength in public cloud overall.
In public cloud, I think the most important point that we drill home is we've had these relationships now for going back. We first put ONTAP on the public cloud 10 years ago. We introduced a first-party cloud-native service first with Azure about five years ago, with AWS a few years ago, plus or minus, and with Google about a year ago. And we're the only provider out there that has a first-party native integrated file storage solution in all three of those major hyperscalers. And that really having that vision across all of them, I think, has really given a boost to our business going forward. Certainly, if you look at the mix within it, and we don't necessarily disclose that, but the mix is obviously weighted towards the service that's been around the longest, right, because it has the longest tail.
We just continue to grow each service. As each of those services come online, I think you've seen that growth trajectory, you know, get much stronger. We also have put just work over the last several years just in the technology and continuing to reduce the cost, improve the performance, and just really address some of the technology fundamentals there that continue to make it even more competitive so that customers, as they gain awareness of these on all the different hyperscalers, they can automatically see the ROI and TCO of moving to those sort of file storage systems. They're just choosing to execute at increasing rates.
Great. Maybe talk a little bit about, you know, the offerings that you have and how you think they're differentiated, and is there risk that over time some of the other storage vendors may try to replicate what NetApp has done? How hard is that to do? You obviously have a, you know, pretty good incumbency advantage here.
I think, you know, and I've alluded to it before. If I had to pick one advantage for, and there's many, but the whole concept of this unified data storage and the leverage we get by being able to use ONTAP for all-flash for performance flash, for capacity flash, for still hybrid flash systems are still a decent part of our business, and then on all the major clouds and use them for file workloads, block workloads, object workloads, it just gives us a tremendous amount of R&D leverage. And our competitors are going to attempt to match that, certainly. But I think there's a pretty large moat in terms of doing that, in terms of how long it takes to develop a truly unified data storage system.
And even the competitors that we've seen have tried have kind of very slowly gotten there. Most of our competitors have grown inorganically, and they may have five or six different storage operating systems. It's very hard to unify a fractured sort of installed base to get there. And it's very hard to, without making compromises, build all of the data security, all the data protection, all the storage efficiency, everything we built into ONTAP over the past 30 years and catch up to that. So it's certainly, you know, the industry isn't standing still. There's competition out there, and I'm certain they're trying to catch up. But we're also continuing to build a forward moat ahead of them.
Every time they catch up with us, we introduce something new, like being the first to really introduce anti-ransomware technology powered by AI directly into enterprise storage to provide real-time 99% plus accurate detection of ransomware. Building those sort of moats on top of things is going to ensure that I think we always have a competitive differentiation.
Great. Maybe if you can go into a little detail about, I would guess the main competitor is actually the native storage offered by the cloud themselves. So maybe just walk us through kind of how customers look at that ROI or business case.
Yep.
to use NetApp ONTAP, instead of the native storage that's offered by the marketplaces.
It's, it's an interesting point. And I think there's a sort of fine distinction because we don't look at them as so much as competitive because the, the interesting thing is each one of those offerings is actually native offering of the hyperscaler, right? So AWS, FSx for NetApp ONTAP is an Amazon offering. Azure NetApp Files is an Azure offering. Sorry. The AWS is obviously an AWS offering. Azure is an Azure offering. And the Google Cloud NetApp Volumes is a Google offering. And so for their sales teams, when they're going out and talking to customers, it's the same product for them. They get compensated just as if it was an Amazon product or so on and so forth. So we're not really competing. From an Amazon perspective, they're happy to sell the storage in whichever category makes sense.
And in fact, if you go to, like, AWS re:Invent, they had tens of breakout sessions where they were featuring. And it wasn't NetApp sessions. It was Amazon sessions featuring FSx for NetApp ONTAP, and some of them don't even say on the tin, "Hey, it's about NetApp." It's just, "Hey, this is the right tool for the job." So a lot of the business gets driven actually by Amazon, Microsoft, or Google towards those products. I will say if you come in with unaided awareness and you're just doing self-service, then yes, you're right. The natural inclination is you're going to go straight to EFS or Azure Files or something like that. That's just, that's natural. But what we find is for customers, they can immediately look at the ROI and the TCO calculators we have on there.
It's incredibly easy because it's a first-party service to put your first terabyte on that and just kick the tires, right? There's no stand-up time. There's no real cost other than your experimentation period. And so a lot of customers will take the workloads they're running on the native cloud storage, and they'll move it to our native cloud storage there. And they'll see the performance improvements. They'll see the cost savings from having storage efficiency. And it's just an instantaneous ROI, and that allows them to move over. So in a lot of ways, it is, as you kind of said, it's about building awareness. And once customers are aware of the solution, in almost all cases, it proves to be a superior solution.
Otherwise, to be quite candidly, none of those partners would want to work with us if our solution didn't offer value over what they had already built.
Right, it's helpful to have that added go-to-market with those three.
Absolutely.
Large players. Maybe just closing this part out, if you could just talk about NetApp's go-to-market. And, you know, you talked about having, you know, specialists in AI. How are you tackling the, you know, the public cloud business, different offerings with, you know, go-to-market or engineering or sales? How do we look at that?
It’s very similar, I would say, to AI, although obviously more mature since we’ve been doing it even longer, going back five years or more with the first-party service. All of our sales teams, and for that matter, all of our channel as well, are trained to be able to sell on-prem and in the cloud. And in fact, I think one of our value propositions is we don’t go into customers with a religion about it. We go into customers with, “What’s the problem you’re trying to solve? What’s the workload you’re trying to solve? Do you want to do it on-prem? Do you have a preferred cloud provider in mind?” And they can open up that conversation. So our entire sales team is trained to do those conversations. It doesn’t have to go through a specialist channel.
When you get down to a decision or when you get down to really investigating, " I want to look at Azure NetApp Files," then we have a cloud specialist team that can come in that is, you know, deep into Azure, right, perhaps even more deep into Azure than deep into NetApp, and can really focus on what's the right way to build out that architecture on Azure NetApp Files, what are the reference architectures we can bring to bear, and how can we build the best solution for that, and so we have sales specialists there as well as technical specialists, so it's a very similar model to sort of the AI model where we get leverage across our entire sales force, but then we have the specialization we can bring in to close the large deals.
Maybe one more on this. I lied. Can you talk a little bit about having, you know, the strength on-prem and what that means to the cloud offering? And, you know, are you seeing a fair amount where, you know, ONTAP is on both sides, or, you know, are you seeing a little bit more, you know, just cloud-type deployments? How does that track?
I don't think we say the mix entirely, but there's definitely a good distinct mix of customers that are on-prem and in the cloud and customers that are just on-prem and customers that are just in the cloud. They're really all of those are true. One of the nice things for us, especially about the cloud business, is especially with all three of these major hyperscalers acting as a sales team as well as they drive awareness of NetApp. They drive awareness of ONTAP. Once you start using something like FSx for NetApp ONTAP or Google Cloud NetApp Volumes, when we go to do that re-compete, right, so say they're incumbent with one of our competitors on-prem, they then have that experience with our software. They know it's a proven quantity.
We've already established the relationship with them, and it makes it easier. So we've actually had sales from on-prem that are really conversions from cloud storage as, as well as vice versa. It's a nice, it's a really nice balance and a really nice approach that we think is really differentiating the industry because no one else has those sort of tools that they can use to go any direction that it makes sense to acquire that customer.
Great. Great. Maybe let's get back to kind of some of the core product areas. All-flash business has been really strong. I know the QLC NAND products have been great. But if you can just talk about kind of what's driving traction in NetApp's business in all-flash, and then we can get into more specifics.
Sure. I think there's a couple of things. One is the capacity flash business. We have our NetApp AFF or all-flash product that, as you said, has been doing extraordinarily well. About 18 months ago, between 18 months and two years ago, we introduced what we call the AFF C-Series for capacity flash, leveraging predominantly QLC media. And that has just taken off with customers. We've said publicly that that's the fastest scoring product in NetApp history. And it's just driven a tremendous amount of growth because it gives customers who traditionally, they couldn't choose hard drives or they didn't want to choose hard drives, but they didn't necessarily need all of the performance and all of the ultra-low latency that performance flash offered. There just wasn't a middle option. It was truly an underserved sort of market segment.
And so there were either pricing pressures on performance flash to drive down, or you just didn't make that sale and still stayed on hard drive base because we're really one of the only, and we were kind of the leader in putting capacity flash into the enterprise storage market, and especially into the unified storage market. There was really no one else doing it in unified storage. And even today, it's a minority of the market that is doing any sort of capacity flash. That's just taken off by gangbusters for us because the value proposition for customers in terms of the footprint, sustainability, and the overall cost savings by going to that is just the ROI and TCO are near immediate for being able to do that. And so it's just absolutely taken off for us.
I think that's been one of the main drivers.
It does seem like there's more focus from some of your peer competitors in this area. So maybe touch on. I'm sure ONTAP's going to have something to do with it. But what will keep NetApp's C-Series differentiated and winning versus peers as they try to get more aggressive in this area?
So, I mean, we've seen peers announce that they're going to come out with capacity flash offerings. And, you know, with the exception of, of, I think, one that's had one for a year or two now, all the rest of them have just said, "Oh, it's coming," or they've had it on one part of their product line. Again, you go back to the traditional, more legacy full-stack providers. And I mean no disrespect to this, but they typically have four or five or six different storage operating systems. And so they're having to roll it out one at a time in terms of each of this. So, like, perhaps the one with the largest overall storage market share only has it on one part of their, one smaller part of their portfolio so far. So we do think there's a moat.
There is technology innovation that's necessary to make capacity flash work. It's not simply stick QLC in a system and it works. Otherwise, like, the industry's not stupid. Everyone would have done it yesterday if it was that simple. So there is actual R&D that has to be done there. So we do think there's a moat there. And then if you go into the future, we've been doing QLC so successfully that, although, you know, I'm not stating we have any intent to do this, further innovations like PLC and different things like that mean that we're already ahead of the game in terms of being able to take advantage of that. And then some of the things I talked about with doing disaggregated storage and other things like that will allow us to further unlock the bandwidth of flash storage.
We really think that there's not only do we have a several-year lead, but we're putting so much R&D into continuing to keep a sustainable moat going forward there.
And you mentioned kind of filling that gap between disk, hybrid, and higher-end flash. What kind of applications and use cases are you seeing in that bucket? And, you know, is there just a very large offer? Are we very early innings of addressing that with the rest of the storage market?
I think that it's really been a wide gamut of workloads. It's basically just about any workload that isn't your true tier zero mission-critical workload. From a technical perspective, you know, our performance flash products, we say less than one millisecond of latency. It can go down to 250 microseconds to 500 microseconds, whereas capacity flash is more like two to four milliseconds. So it really is about what are the performance and response time requirements of your application. But I think the vast swath of the mainstream enterprise applications, things like virtualization, containerization, databases, maybe not your absolutely mission-critical tier zero, but all of your mainstream databases, almost all unstructured data that's out there, all of that is a perfect fit for capacity flash.
Obviously, with the massive growth, you know, the zettabyte scale growth overall in the industry in unstructured data, right, capacity flash has a really interesting future ahead of it there.
Right. I wanted to touch on Keystone a little bit. There's always a lot of talk for, you know, companies like NetApp and others in the, you know, enterprise world about more of an as-a-service type offering. Keystone's still small, but it's been doing very well. Maybe, you know, talk a little bit about how customers view that and what's kind of driving the recent traction in Keystone.
S o I think we continue to pick up customers in Keystone every quarter. There's strong interest there as you indicated. I think for us, it's the ability to have a complete as-a-service portfolio from the cloud storage to the on-prem and offer all of that as storage as a service has really brought in customers. Obviously, customers who start with us in cloud storage as a service. They're very interested in doing as-a-service when they go to do on-prem as well. So I think that helps.
The other thing, you know, capacity flash and everything being able to drive the price points down for certain workloads is helping in Keystone as well as, you know, realistically, we just don't drive to customers and say, "You've got to move to as-a-service." So we have a very healthy mix there, and we don't try to be prescriptive. We're really trying to provide the storage to customers as they need it. And so we think that by providing those different options to customers, it really has provided sort of a healthy growth. And one thing we said at our user conference was that of the Keystone customers we've acquired, we've had a 0% attrition rate so far. So that's very important to us in building that as a healthy, sustainable business.
It's not a sort of flash in the pan, so to speak, or trying to just grow that business as a financial exercise at all costs. It's really about building as a sustainable business when customers want to do it, and we've seen customers, even in the same customer, move back and forth depending upon macro trends between OpEx and CapEx, and we're able to fulfill both of those, and we have a lot of customers that may be Keystone customers and have traditional CapEx and may be moving between them, so it really is a nice full portfolio.
Right. Right. Zero attrition's good. Maybe you should be charging a little more for something.
No comment.
All right. And then maybe, you know, the last one on Keystone, talk about, you know, kind of new logos or, you know, you did mention a fair amount. You know, it sounds like there's some that are starting in the public cloud and using Keystone on-prem. But how is this working as a tool, you know, for the broader NetApp?
I mean, Keystone is a tool that's available to all of broad NetApp, right? So it's not like we have a large specialist team for Keystone. We do have a specialist team for Keystone, but Keystone is so fundamental to how we're able to deliver the service that it's just sold by our mainstream go-to-market, you know, account representatives and account technology specialists, as an option out there for customers. And it's a very straightforward option where you choose your tier of service. So we can give them options for doing it as CapEx for whatever period of depreciation they want to set up. And we can give them side by side with that options for, you know, here's exactly what it would look like as a service.
Here's what your monthly cost would be or your quarterly cost and provide it to them. So it really fits very nicely into our core go-to-market model.
Great. And then I did want to, you know, get back to ONTAP. We've touched on it throughout some of these answers here. But give us a sense of the power of ONTAP and what that means. I mean, it's obviously gives you a very strong ability to keep existing customers. But, you know, talk about the differentiation and how that is a tool for NetApp to continue to outperform the storage space.
So I, I think, you know, ONTAP has proven to have such a durable advantage, and we're able to continue to drive such innovation in that it really is, you know, amazing in terms of how much it's able to achieve across file, object, and block now, with customers as well as on-prem and in the cloud. And we continue to add functionality into it, right? So data protection functionality has been sort of our bread and butter for decades. Now adding, I mentioned the anti-ransomware capability that has driven, you know, we put a tagline out there that we are the most secure storage on the planet. And we stand behind that, for example, because we're the only storage that has been certified to store top secret data, and the only enterprise storage that's been done so by the U.S. government.
We were the first and one of the only to embed anti-ransomware technology into it. We have built-in compliance technology into it, right, that's been certified for SEC-compliant data. All of these capabilities that we built over the decades that are built into ONTAP that are available essentially to any workload, it just, it's a force multiplier. We don't have to have a specialist group for selling to this workload and a specialist group for this storage system and a specialist group for this storage system. It really is, you know, this amazing Swiss Army knife. And the cool thing is we've been able to take that and differentiate over time. So we do have a file storage solution in Azure.
We do have an all-block solution in NetApp ASA that's been doing remarkably well on-prem where we've taken ONTAP and just really focused it into the block spectrum, and we've been able to take ONTAP and use it to bring down all these amazing AI deals we talk about, and we're able to flex it in the future with the vision I talked about to build it into a disaggregated architecture and scale it up to even higher heights, so it's just really, in a lot of ways, the gift that keeps on giving. It's an incredibly flexible architecture under the covers. I'm a geek by trade, right?
And yes, I work for the company, but I wouldn't trade ONTAP for any other storage OS on the planet just because of how extensible it's been to allow us to capture every one of these major market transitions, whether it's virtualization or containerization, now AI. Each one of these waves that's come along, we've been able to catch that wave instead of having it pass us by. And that's relatively unique in the storage industry, right? Everyone else has had to switch storage OSs. They've had to; some companies have just gone under because of these waves. We, each of these waves has just driven us further forward because of how adaptable it is.
Right. Right. Great. I think we're coming up on time here, but appreciate the timeline going on with Keystone AI Cloud. A lot of good things going on for NetApp. So appreciate the time.
Thank you very much.
Thank you.
Thank you.