 Welcome back everyone. Live coverage of theCUBE here in Las Vegas for VMware Explorer. I'm John Furrier, host of theCUBE. We're on the green set here. We've got the blue set over there, Dave Vellante, Lisa Martin, Rob Streche. Great coverage, wall-to-wall. We're on day two. We're talking to all the experts. Top story here is multi-cloud, super-cloud, next-gen applications, of course. Genitive AI, again, is big centerpiece of the show. That's going to increase the importance of data, the importance of the new next-generation architectures, and we're going to unpack that conversation here during this segment with two experts and friends of theCUBE's deep knowledge, domain expertise in storage, networking, and applications. Eric Herzog, famous CUBE alumni, who's the CMO of Infinitad. Eric, great to see you. Thanks for coming on. Thank you, love to be on theCUBE, John. And everyone knows David Nicholson here on theCUBE, also the CTO in the field for talking to customers, bringing it down, both very technical. We're going to get down in the weeds, but also go back up and look at the forest and the trees in storage and data. Guys, thanks for coming on theCUBE. Thanks for having us. All right, so first, Dave, we did a lot of analyst sessions too with theCUBE. Eric, you've been on, we've had so many conversations, Eric, on storage. You're like a walking data sheet. You can pull stats out like that from five years ago, talk speeds and feeds. IEO and data right now, and LLMs and AI are kind of the top story, going fast, doing training. It's the role of storage, storage architecture, storage in the cloud, all part of it. The industry right now is probably at an inflection point where storage will probably be more of a platform than ever before, but the game is still the same. You've still got to store shit. Yes. So this is a really interesting time in storage again on this next wave. How do you see this? You've seen many ways of innovation. What's your take on the landscape right now? Because storage is not just about the hardware, it's about the software, the architecture, how it all fits in. What's your view? So I think there's a couple of key things. First of all, it's now going to be all about availability. If you're going to use AI in your applications and workloads, it can't go down. So the AI is going to feed in to look at the apps. Some people will embed AI in the apps, others will use AI to examine the apps or examine the data so you can't go down. So that reliability availability, which by the way has been a watchword in the enterprise already, is just going to get more important. Performance, obviously. But again, that's also been a watchword in the enterprise before. So I hate to say it. It's almost like some of this stuff is coming full circle. It's just that AI is exacerbating it. And then obviously the two new trends are one, because of the up and down economy, how does storage impact your return on investment TCO? Because it's expensive stuff. So how can you cut costs? CIOs are never storage guys. They don't really like to spend money anyway. And then the last thing is really how does cybersecurity fit in? As you've seen all the AI experts talk about, well, if ever the cyber moth gets ahold of AI, we're in bad shape. And guess what? So the guys who are fighting it off better have cyber on their side. Yeah, and I was talking to a VC, kind of a later stage VC. I asked him about SuperCloud and securities. Like, at the end of the day, John, it's about the cheddar, the cheese, the money. Data is the cheese and the mouse wants the cheese. So you got to protect the cheese. And protection is not just data protection. And ransomware, which you will talk about, but with AI, a lot of conversation around private AI is here at the show at VMware, meaning the data's the IP. So this idea of the role of data continues to evolve where it's not just store and move from A to B. It's about the value of the data in all phases of its state. This is a huge, Dave, this is now another dimension. It's almost like 3D chess and storage with data. It's going to be valuable. And the hell I keep it on-prem, when I'm moving to the cloud. So there's policy around data for other purposes besides security and disaster recovery. It's like more maybe revenue driving. So kind of new waves are coming, new ways, new ways of thinking. How are you guys seeing this from a customer standpoint, Dave? What's your take on that? So I absolutely agree data is the new cheese. I've been sitting here waiting to say that. I think what's interesting is, as these evolutions occur, Eric just mentioned, it's sort of cyclical. And I saw the announcements around VSAN Max. And it was interesting to me when it was first described as disaggregated, I thought, well, okay, that must mean disaggregated in terms of location physically, but it's like, no, no, no. The term disaggregated is used to denote disaggregating the storage from the server. In other words, unconverging hyper-converged infrastructure. So you could call it disaggregated or you could call it unconverged aggregated storage. Hyper-unconverged. Hyper-unconverged storage. So, but it makes sense. We had this issue when we went to the hyper-converged model of this thing that I'd refer to as the devil's triangle of CPU memory and storage. And if you said you needed new storage, more storage in that model, you had to buy another server. And so like the stuff we do in Infinidad is designed to, no, no, no, no, aggregate all the storage and share it out. That's not a new concept. But so it is interesting to see the next iteration of VSAN going in that direction. But no, it's always been about data. It always will be about data. And all of these tools are about surrounding the cheese like eager mice. Yeah, I mean, everyone wants to get the data. The data is a killer part of it. Eric, Infinidad's approach to the market relative to some of these modern discussions, modern applications are coming, this abstraction layer, super cloud, multi cloud environments, which today are essentially customers with stuff in clouds, from multiple clouds in one place, like toys sitting around, they maybe work individually, but not together. There's a movement to make this stuff work together. How are you guys seeing this from a storage standpoint? Because end of the day, storage is a social truth at the end of the day, that's where the data is. So we see a couple trends. First of all, we see seamless hybrid cloud integration. You've got to be able to move data back and forth between on-prem and off-prem. Second thing we see is because the value of the data, and even especially because of the AI, there's issues of how do I keep it protected, in this case, legally protected. As you know, some of the things with AI, people have done it, and you could just do it yourself. And what do you find out? Oh, that's pulling in patent data, copyright data, data that violates GDPR and other privacy laws. So in fact, several major Fortune 500s have told their, even in high tech companies, have told their employees, do not use this because they're worried about getting sued by people who own patents and copyrights and everything else. So the whole idea around AI actually is partially going to push certain amounts of data back on-prem out of the cloud, and then they're going to try to wall it off because the last thing they want is someone's working on their code, they use AI. Next thing you know, oh wait, it's open source now. I just had 500 engineers work on that. I was spending $5 million a year, and wait, now it's open source. I'm in bad trade, public, you know? It's interesting, last year, I believe you said this kind of same concept last year, and what's interesting is it's expanded to the point where here at VMware Explorer, they had their general counsel come out on the keynote. Kind of gimmicky to make a point. I mean, all she's basically saying is like a showpiece. But the point is, that's the level of legality issues involved in this kind of like collision between data and kind of the hype of AI and the hope of something new. I mean, it's got to be reined in. So I would argue data is way more than the cheese. Okay, data is like the gold, the diamonds, and the platinum of the company, not just the cheese. If you find those things more valuable than cheese, I will grant you that. So it's all about, you know, storage is really about applications, workloads in use case. That's the other thing, a lot of storage people talk about speeds and feeds, and no CIO really cares about that. It's, can I make this AI workload that take three hours, run in 20 minutes? Oh, okay, that matters. Can I take a traditional workload, Oracle or SAP, that runs in three or four hours? Can I make that run in 20 minutes? Can I make sure that that Oracle workload never goes down? I don't think about, again, the guys who run IT are all software guys, they're all CIOs. They don't, all they know is that it's sitting on this platform, and that thing better not fail, better not go down, and boy, it better be fast. And so, again, it's making it full circle. In fact, all AI is doing is reinforcing what had been the 30-year-old IT enterprise thing, and now it's just giving it new life. So it's very good. Explain this, this is important, because I think this is the key point. The game is still the same. It's just different environment. What's different? Is it the speed of the change, performance systems, the pressure of the business? It's a pressure on the business to leverage these new workloads. Whether that be an AI-based workload, a container workload, whatever type of workload they're trying to do. And the issue is, certain of those things require even more performance. Yeah, I've been doing storage for 45 years. So I'm an old guy, right? I'm almost 70, and I'm telling you, I've never, ever met a storage guy, or a CIO, or anyone in between, and you said, hey, we can make that application run three times faster. They said, oh, we don't want it faster. And of course, you have people say, oh, we never need it that way. And then you turn around, and when you're in an account call, they always want it faster. They always want it always available. And AI exacerbates that. Hybrid cloud exacerbates that. And with data all over the place, both in the cloud and outside the cloud, you can't know what your data's up, and it's not working. It's got to talk to the cloud, and the data on-prem's down, or the cloud data's down, and the on-prem is looking for it. You just lost those workloads. They're not going to work anymore, so it's very, very critical. Dave, I want to get your perspective on what he's talking about, because you also, this field CTO of Infinidat, you're also involved in CTO communities where you're talking to folks. What's the mindset right now of technical architects and CISOs and CIOs as they come through the psychology? Look, I have a lot of pressure. I want things to run fast, and maybe I might want to have stuff run slower if it's for policy reasons, but for the most part, I want my apps running great. I want my business to be booming. It's going to be digitally native. What are some of the psychology thoughts going on in the minds of these people? Because there's almost a kind of a green field and brown field opportunity going on at the same time, kind of coming together. Do you have any insight in color you'd like to share around what's going on in their mind? Obviously, you talk to them. There's probably projects to be worked on, deals to be done, but what's going on in their mind? Work backwards from that customer and share your opinion. So a lot of it has to do with coming to a real understanding of what the infrastructure requirements are, coupling that with what are we going to do with this AI thing? And a lot of the projects that I see underway are things that aren't going to be the sexy hallmark of the dawn of the age of AI. They're going to be optimizations of existing business processes. But the big question is, against the backdrop of something like VMware Explorer, this question of, okay, well, what does it mean? What does AI at the edge really mean? What does it mean to be able to have a virtual machine with 16 NVIDIA GPU cores contributing to the horsepower for that virtual machine? What exactly does that mean? And what are the fundamental requirements behind that? We just saw, you know, VSAN Max talking about disaggregating. Well, why? You would think that the sort of knee-jerk reaction from infrastructure perspective is, no, no, no, what you want are these nodes with computing storage in them. That's going to be the answer for it. But that isn't always the answer. And there are different phases to AI in terms of whether you're talking about queries that are leveraging language models. The actual inference engines themselves, what they do to create and learn, we don't talk about machine learning as much anymore, but really it's part and parcel with AI. You're teaching the machines and you're asking the machines, how did you get that answer? And then saying, and then telling the machine, that was wrong, don't do that again. That iterative process. So the short answer is they're all freaking out trying to find prompt engineers because they need people to combine whatever domain expertise they have with the technical know-how to leverage these things. And in the middle is infrastructure. And what we do is try to provide some logic behind the scenes to be able to address what their requirements are from an infrastructure perspective. But that's, it's a lot of confusion now. These are early days, a lot of hype. Look, senior leaders that I work with, typically this isn't their first rodeo. And so what they're trying to sort through is they're like, okay, what is this like? I know this is like something I've done before. What is that thing that we've experienced before that this is like? And there are a lot of analogies to be had. It's not, it's not something completely new. And so this is why we think there's a lot of architecture going on and zooming out, trying to look at the landscape. But day to day, they got to make decisions. Where is infinite fitting in? How do you guys vector into these conversations? Because I can imagine you have the tactical conversations. Hey, we should be buying our stuff. We have value to provide, get that. What's the cherry on top? I mean, you've been 40 something years, you know, storage like anyone else. What's different and what's the same? So what's the same is what the underlying infrastructure can do. It varies vendor to vendor, right? Some people can really do cyber resiliency and some other people put it on their PowerPoints and they can't do Jack Poo-Poo with it. And then it's what's the real value? And I think the thing that storage people are realizing not just us, but our competitors. It's about applications, workloads and use cases. How do you impact them? How do you protect them? How do you enable them? How do you make them faster? How do you make them more secure? And the one thing that's happened is prior to the chat bot and AI taking off literary in the last couple months, go back six months ago, it was all about security, security, security. I mean, for a couple of years about IT security. The number two concern of CEOs in the Fortune 500 2023 survey, the number two concern at 22% was cyber security. The number one concern at 26% was the recession. So cyber isn't going away. And AI was all the buzz price since January, but you go back to, it's all about cyber. And I'm telling you, if you don't have the right security around AI in particular, you're really in trouble, independent of the legal and compliance issues, which make it even worse. People are going to try to steal the stuff. They're going to try to use AI, by the way, to steal it from you, which is going to make it harder to detect. So all these things are wrapping together, but it's really application workload and use case. And what can you do to optimize those things for me? Performance availability, security. What are you going to do? And by the way, it needs to be seamless because my guys, I need to spend money on AI. I can't spend money on storage. So how does your storage, if you will, sit in the corner and just automatically do stuff so I can hire less storage guys and hire more AI software engineering talent? That's what you got to be able to tell them. So you got the AI has to be part of it, maybe it bolt on and integrate it. Well, we use AI inside of our box as we call it autonomous automation. We have probably 40 to 50 references that are public where they say we've had our infinite product and we haven't touched it for two to three years. Now, how are they not touching it? All right. Yes, it's rock solid, but it does it on its own. AI built in, if you will. Explain why that's going to be hot because basically what you just said was you're using AI to make it operate on its own, basically. Right. And you get a lot of success customers. Why are they successful? So the key thing it does is allows them to cut their cappecas and op-ex and use that instead of on infrastructure on AI software development or AI integration with Oracle or whatever their workloads are. And there's only so much money. So if you can cut money on storage and make it autonomously automated, doing its own thing, but still functioning the right way, then you can spend the money on these other projects, which are the quote, cool projects. And there is a limited budget. They're not just running in the CEO and saying, yeah, oh, you want to fund AI, Mr. CEO? Sure, here's $12 million. He's like, no, figure out how to do it with the current budget I gave you. So VMware had a lot of announcements. They had like the cloud orchestrator, NSX Plus, ransomware, desk recovery. They had, you mentioned VSAN, Max. Is that the right direction? Or, I mean, it sounds like you said it's unconverged, which is kind of tongue-in-cheek. Hyperconverges was the big wave, right? Is that a good thing or a bad thing? Are they going down the right road? Is that different from what Infinitab's doing? How would someone make sense of that? No, I think, no, they're simply with VSAN, Max, they're simply acknowledging the fact that sometimes having those units of scale that are restricted in terms of you must buy CPU if you want more storage. If you, oh, but I really just need more memory. Okay, well, you must buy CPU and more storage. The ability to scale independently just makes sense. So that makes sense. Now, the question is, objectively, from an industry perspective, will something like disaggregated VSAN, Max, which looks and feels a lot like what we would call a storage array, in a sense, just with the Lego pieces put together in a different way, what will it offer in terms of costs for performance, cyber resiliency, and all of those metrics? That'll be a question moving forward, but in a broader sense, it will be really interesting to see how Broadcom's continued investment in VMware, as it becomes part of Broadcom, will play out in this hybrid cloud space. We have the ability to allow certain kinds of workloads to interact between on-prem and the cloud. VMware, for years, has been talking about this software-defined data center stack that includes network, CPU, and storage virtualization. Everybody loves the CPU virtualization. People have been kind of so-so on the storage virtualization, and frankly, the network virtualization hasn't been as popular as maybe people here would believe. So it'll be very interesting to see if you must have NSX deployed on-prem to leverage some of these things. You know what, we're okay with that. The thing I love about being part of Infinidad is we try to be the best partner to the dance. We don't pretend to be the center of the universe. Some of the stuff we were demoing here has to do with supporting cloud, what they refer to as cloud-native storage, for Tanzu environments in VMware. We're fine, we're fine playing that role. The true test of VMware, in my opinion, is going to be can they handle the truth when it comes to NSX, and will they give customers the option not to have it a requirement or dependency of something else? That's going to be interesting, because I don't think a lot of their customers are using NSX relative to the overall percentage. So we're going to, I'm going to keep an eye on that great, great call out. And the final minute we have left, guys, and first of all, thanks for coming on, Eric. Great to see you, Dave, you're awesome. Great expert advice here. Give a minute or so to give a plug for the company, Infinidat, what you guys are up to. You're the CMO as well as you're kind of like a product czar, too, practically product and CMO. What's the main value, but why are you winning? What's the pitch? What's going on with the company? Give a quick plug. So the two things that are having us win is around how we help optimize real-world applications seamlessly because of the automation. B, how we help them cut costs. Okay, we can consolidate. We have a customer, 25 different storage arrays. They have four of ours. They saved 10 million. We have another one that had 120 from another company. They now have 44. They've saved 60 million in the last two years so they can spend it on AI. So that's the value. And then lastly, the cybersecurity thing. We've imbued our storage with all kinds of cybersecurity technology. We can recover 30 petabytes of backup data set in 12 minutes and we've done it live 10 times. And we can recover 200 terabytes of primary storage data in four seconds and we guarantee all that in writing. So it's really about doing SLAs and being about workloads and not being about storage speeds and feeds. Trust me, I can tell you all about a hard drive. I used to be at IBM and Mac store. No one cares, right? I can go back and play the tape on the queue. I got fresh Eric Herzog's data sheet speeds and feeds button. You go on forever. You energize our bunny. I know you know your stuff on the speeds. That's a good point on the customers because if they can get the cost savings, okay? Right sizing's been called about. Been talked about in the industry. Refactoring, right sizing has a better word. Just clean up the footprint, not lose anything but gain and save. That's cash for AI. Which by the way, which by the way is also under scrutiny for unpredictable costs as well. So I said a few pennies here but lost a bunch of Benjamins over here because I didn't understand how training works. And we're seeing a lot of that going on where the optimization of cash can really get a blind spot on where the investment should go. You seeing that same thing? Again, people want to move five years ago. Everyone want to move to cloud. That was like the hot thing. Not everyone wants to do AI, it's the hot thing. So what that means is an infrastructure provider, not just me but the network guys and server guys, show them how to save dough, show them how the apps work so they can spend that money now on AI. It used to be let's spend the money on the cloud. It is the repeat. It's just that instead of cloud, let's spend it on AI. And five years from now, it's going to be let's spend money on Henkelmeyer, whatever that is. Super cloud's going to be it. Eric, Dave, we went over as usual two great experts on theCUBE here. Dave Nichols and Eric Hughes are with Infinidad. I'm John Furrier, host of theCUBE. We're back with more day two coverage. We're going to five o'clock, 5.30 up until the very end, until they kick us out. We'll be back with more live coverage after this short break.