Transcript
Hi Mike Matchett with Small World Big Data and we are here talking with Infinidat today. Infinidat has some new releases in which they've really improved tremendously the capacity and performance of what they're doing. We'll get to that in a minute. But first we're going to talk with Eric Herzog about, uh, some of the things that are going to happen in 2026, some of the things they see coming, some of the things we're going to want to react to and be aware of. So just hold on. Mike, thank you very much. We love working with Small World Big Data and 2025 was a great year in the storage industry. 2026 is looking to be even more exciting. All right, so Infinidat is sitting at the edge of enterprise high end storage. But as we were just talking before offline, uh, it's it's not really about storage ever anymore. We're talking about when you're doing things like this, you're investing in high end storage. You're talking also about cyber security. You're talking about empowering other workloads. You're talking about getting more power efficient in your data centers. So let's talk about some of the things coming in 2026. Um, let's start with, uh, something coming up. Rsa is coming up RSAC. I should say, uh, tell us about what you see happening in the cyber storage space. What is cyber storage resilience and what role do arrays anymore have to do with cyber storage? Well, basically when you think about the traditional cyber storage security, whether it be driven by a CIO or a CISO or CTO in general, they've ignored storage. Yet 90% or more of a global enterprise is sitting on their enterprise storage. So the cyber criminals are not dumb. This is not like the old gangster movie, right? Where you see them driving in that old car with the machine guns out the window. These guys are technically sophisticated. They know where the data is. They know it's on both primary or secondary. So first of all, in 2026, even more CISOs, CIOs and CTOs need to recognize that enterprise cyber security is a critical component of a comprehensive enterprise cyber security strategy. That's the big thing in 2026. People need to understand that's where it's going. Now we embed inside of our technology and inside of our family what we call Infinidat. It's won over 20 awards not only from storage analysts such as yourself, Mike, but also from the cybersecurity analysts as well. So we've hit a real, real strong chord with Infinidat. And we would argue when you take a look at the vulnerability of your data, it's not just at the edge, the perimeter, right, if you will, the moat with the alligators and the piranha, and then, of course, your castle wall. Last I remember, Robin Hood, who was a good criminal. Not like the cyber. Criminals. Right? The good guy. Remember, Robin Hood got over the moat and over the wall, and they were trying to track him down. So. You don't want to do that. And the question isn't if you'll be attacked. It's when and how often. In fact, the average global enterprise suffers over 1650 attacks per year. And that number is going up year by year. So if you put storage in to cyber security strategy, you're covering everything from the edge, your servers, your applications, your networking. And now where the data is sitting, not where it's transporting, right. The edge, the servers, the network that's the transporting right and the compute side. But the data ends up on storage, both primary and secondary replication archiving. So you need to protect that too. And that's what the CIOs and CISOs need to realize. I mean, I definitely think that people outside the storage industry, first think of protecting storage. As you know, I've got encrypted, encrypted data on disks which we know now has to go way beyond that. We need immutability, but we also need these rock solid like you mentioned in passing secondary storage that's tied closely to the primary storage. You can't you can't just make backups and assume now you're safe because those backups are also vulnerable. Right? The whole thing has to be inside of a larger strategy. Uh, and that's definitely where I think Infinidat has some great solutions. Uh, as you mentioned. Um, let me just talk about another thing that I heard at, uh, last year. Well, this year in 2025, coming into 2026 was AI. Of course, AI is everywhere. But we've got, you know, Robin Hood's wearing a green hat. Maybe, but we've got we've got we've got AI for good and AI for evil. And it's it's a huge thing. This is just across topics here. Then, uh, what do you, what do you see coming in in the uses of AI in cybersecurity. And then, then we can talk about it as a general workload by the good guys. Right? Right. So one of the key things that we already use AI and ML technology and what we called infinite safe cyber detection. It's designed to scan VMware data stores, volumes, files, databases, snapshots, full arrays only a volume of an array, only 12 snapshots in array. But the idea is you use infinite Safe Survey detection, which uses AI and ML technology to search for cyber anomalies, particularly ransomware and malware. It can be integrated and communicate back to your SIM or your source software, or if you have a security operations center. And that way you can see we recommend customers check on a regular basis. So when you think about things like SIM, saw and SOC on the storage side, they're the early warning system, right? They're sensing there's a cyber attack. If you do this proactive testing right. And you set up to go once a week, once a day, whatever makes sense, then the storage can be an early warning system. The second use case is clearly you've had an attack, so Small World Big Data has had an attack. You know you've had an attack. What? To recover. You need a known good copy, right? Yep. So malware and ransomware are not King Kong and Godzilla. They are surreptitious. They are the ultimate James Bourne. So you could be taking immutable snapshots or backing up malware. Ransomware. So with Infinite Safe, we create a fenced forensic environment. And then you could use the cyber detection capability to analyze potential candidates to recover. So that's using AI and ML in the cyber world for good. All right. All right. So let's let's I mean there's probably plenty to dive into there. It'd be great to even demo some of this, but uh, let's talk now in 2026. Going forward, AI is a workload that everyone's trying to run and not having enough infrastructure in many dimensions to do that. How's how's how's Infinidat positioning? Uh, what you guys are doing as as in helping people empower, empower themselves up with AI. So we see a couple areas there. We see AI as an application of workload and AI also for use inside of your storage system to make it easier to manage and more cost effective. We do both. So we have an AI solution. This is a Rag reference architecture. Um, it works with Infinidat storage. It'll work with our competitor storage as long as the data set, whether that be block with as you know, all the databases now support NFS or whether it be NFS file data. If you want to use us with a third party vendor and you've got us in that third party, as long as the IT guys know where the databases are with NFS or the files they want to bring into the vector database and ultimately run through their LM or SLM, we can do that. Okay. So that AI rag allows you use existing technology. The idea, of course, is do it in real time. Data changes all the time. And of course, as many of the listeners are aware, AI, if you do it open market aka not inside your own data center or inside your own private cloud can expose you to all kinds of legal and regulatory issues, right? If you're publicly traded company and you're using open AI, right? Something open sourced and you start using it on your financial data, you just made that financial data public, which of course, for publicly traded companies, except when you at the end of every quarter and every year is illegal. So if you're doing it internally on your own storage or on your own private cloud, that way a the data is yours, you don't have to worry about that. B by using the AI rag you're updating in real time if you're using it, for example, for supply chain and you're building rocket ships or cars or planes, you're building stuff every day and you've got factories all over the world. If you're a global Fortune 500, so you've got to make sure you have the right but bolts, the right nuts, the right screws, the right windows, the right whatever it is you're building, you need all the right stuff and you have to ship it all over the world. So if you're doing it for in that way, you need real time inventory information. What's in the pipeline, what's shipped already from your vendor. And that can be done with AI helping write your logistics team and your supply chain team do their job, but you don't want to do that open. Because guess what? Now everybody knows what you're ordering and everybody can figure out, oh, they're building this many cars this year. Okay, I got it right. So you want to do that privately and you want to do it constantly in real time because it's changing every day. So our AI helps you do that. So that's how we see it. And by the way, don't forget AI is just another workload. Maybe AI. It's got all the sizzle now. But guess what? So did early relational databases have all this. Oh don't don't just don't don't don't put age on us. We both got gray hair now, but don't, uh, let me let me say say you've got a hybrid version out now in in one rack that can hold, what, 33PB. 33pb, effective in a single rack. Single rack. So when we're talking about rag and we're talking about rag at scale, and we're talking also about the fact you've also got highly efficient replication going on now. So I can get that data around the world as well as in there. Uh, it's not just like, you know, here's 100 gigabyte database. I need to rag or a file, one little file volume. It's massive amounts of data we're talking about making available to the AI and, um, and performant. Right. There's a whole point of Infinidat if I, if I understand. Right. Um, so, uh, but there is a cost to this, and it's a power cost usually. And data centers are sort of running this high end storage, all those GPUs, all those nodes, all those clusters, it starts to look like an HPC design that's just sucking a nuclear power plant dry. Uh, what What are you guys doing at Infinidat to address power this year? So we're all about power efficiency. We do it from the design up when we design. So for example, our initial G4 platform, which was announced in May 2024, we made it 70% more power efficient than our older generation. Okay. Now at the end of September 2025, we did a powerful evolution and refactoring of that, and we brought out a new high end enterprise array that's only 11 Raku. So by leveraging that, you still get our 35 mix of latency. You still get our cybersecurity within safe. You still get our benefits of leveraging AI inside of our array with what we call Infinidat that helps you more easily manage and proactively, for example, tells you how much power you're using, how much capacity you're using, and you set that all up proactively. All great stuff, right? Well, you've got to make sure that that's power efficient. And with our new 11 rack, you. You're even that much more power efficient. In fact, when you look at some of our competitors out there, compared to one of them, where 6.9 times more power efficient model to model, same type of model, right high end enterprise 6.9 times and in fact, compared to many of the upper mid-range products were actually equal in power efficiency. But we offer 100% availability guaranteed. We offer cyber resilience and RTO guaranteed. We offer performance guarantees, and those mid-range solutions don't. Yet our power efficiency the same. We view it in a couple of ways a it can be a greener environment. B you could save money if you want to, but most likely, most likely, what we see is that customers are going to take what they're using in power for storage and put it towards AI or other workloads, because whether it's an AI data center, as you mentioned, or even traditional data centers, they're thirsting for power. So all of that is what you need. So we think you could save money with the power efficiency. If you want to, you can use it for other workloads and use cases and of course screen it. So we see power efficiency as a critical factor in how we design and deliver solutions to our customers. All right. Let me ask you this. You mentioned the fact you've got a smaller form factor offering from filled with high end storage. So it's still a high end storage design. But are you are you trying to now convince what might have been a former midrange kind of storage buyer to buy high end storage, or are you offering something new to the high end storage customer? How does that really fit into the portfolio? And yeah, so a little bit of both at the very, very upper reaches of the midrange there, they're almost expensive. And we've dropped our entry price point with the 11 Raicu 77TB. So if you need 77TB and you're going to buy an upper mid-range product, our solution has more feature function benefit at basically the same price when you go lower down the totem pole. We don't have solutions there. So at the very upper mid-range, there's a little overlap with the high end. And that's where we can, um, use that. And customers can use that if they're going to look at an upper mid-range solution. Still, you're pulling the market away from this idea that storage is ephemeral and should be done in a container on some device that no one knows where it is to something that says, hey, we've got something more power efficient, denser, uh, more secure, far more secure for this, uh, and you're bringing that into the market down? I would say I hate to say down, but you know what I'm saying? You're compressing the market. Yeah, absolutely. And by the way, it's a hybrid multi-cloud world, so we can seamlessly traverse out to the cloud. Um, that said, the bulk of our customers Global Fortune 500, we have 26% of the fortune 50. So most of those customers run a hybrid multi-cloud environment where the bulk of their data is in a private cloud. And then we need to traverse for replication, for backup, for archiving, or for burst. They'll go out to the cloud. But the bulk of their data and what they're doing in these high end enterprises is still being done on prem, but it's not really on prem. It's in a private cloud that private cloud mimics a public cloud and can communicate with it with things like our cloud edition. So very straightforward way to have a hybrid multi-cloud environment. So not all data is going to be on prem, but not all data is going to be in the public cloud either. It doesn't doesn't mean if you were doing a bunch of your AI all the time in private, in public cloud, you just exposed your financial data if you're publicly traded. So you can't do it that way. You got to have a mix between off prem and, if you will, off prem public cloud and on prem private cloud. And that's what our global fortune 500 enterprise customers. Do especially I mean, if you're if you're cloud building but you're not one of the hyperscalers, right. You you've you've got to have a solution for the storage side of things that has all these guarantees, all these privileges, all these, uh, efficiencies to it. Uh, but it's dead simple to operate because you're doing it as a service, right? So, um, you know, I think that's definitely a growing business area for Infinidat is, uh, is that kind of private cloud building mentality in architecture, right? Absolutely. Absolutely. Mike. Yeah. Uh, so so, Eric, I, you know, there's lots of things to talk about here, uh, with your with your with your latest, uh, releases. Lots of things coming 2026 roadmap. But we're kind of out of time for now. Maybe we'll catch up at RSAC. Yeah. No problem. But, uh, if someone wants some more information right now, they see this and they say, you know what? This is probably the year we should step up and look at some real storage. Uh, what would you have them start their research? So one, uh, WW infinidat.com. We also have a following on LinkedIn. You can reach us there. And also of course we still have a Twitter account. So at and I. Think it's called X now. Not not to not to not to give it away. I'm, I'm an old I'm an old geezer, as you know, Mike much older than you. So it's still Twitter. Yeah. On there. But definitely check out Infinidat. Um, lots going on there. High end storage is not the dinosaurs. You guys think it might have become some of you? I've heard that it is actually progressing and evolving and becoming something almost agile in a way. Uh, high end storage. Imagine that. Great. Well thank you Mike. We love working with Small World Big Data. And, uh, again, 2026 will be exciting for enterprise customers. 2025 was 2025. 2026, I think is going to be even more exciting than 2025. Was. Uh, exciting? Hopefully in a good way. So but take care. Thanks for being here. Thank you for watching. Check out Infinidat guys. Take care.
