RAISE Summit 2025 | Rajiv Ramaswami, Nutanix

Clips
News
More from RAISE Summit 2025

Rajiv Ramaswami

CEO

Nutanix

12 signs AI infrastructure is reshaping the IPO market: theCUBE’s RAISE Summit insights

Capital markets are being reshaped by artificial intelligence, with AI infrastructure emerging as a critical pillar of how companies scale, compete and go public. What was once considered a back-end concern is now front and center in IPO roadshows and investor decks.TheCUBE’s John Furrier talks about AI infrastructure at RAISE Summit.The shift is about more than hype. Investors are scrutinizing how companies embed AI into their business models — not just to drive innovation, but to ensure long-term operational resilience. It’s no longer enough to show top-line growth; firms need to demonstrate that their AI capabilities can sustain profitability and differentiate them in crowded markets

play_circle_outline Coverage of RAISE Summit 2025 featuring Nutanix CEO Rajiv Ramaswamy.

play_circle_outline Transforming Enterprise AI: The Role of Turnkey Solutions and Accessible Infrastructure in Simplifying Generative Applications

play_circle_outline Nutanix and NVIDIA partnership for AI infrastructure and solutions.

Info
Transcript

Rajiv Ramaswami, Nutanix

Rajiv Ramaswami

CEO Nutanix

In this RAISE Summit interview, Nutanix chief executive, Rajiv Ramaswami, joins theCUBE’s John Furrier to unpack how Nutanix and NVIDIA are building a turnkey stack that brings generative AI from proof-of-concept to production. Ramaswami explains how Nutanix’s invisible infrastructure, combined with Kubernetes and the NVIDIA AI Enterprise suite, creates shared inference endpoints that hide complexity and protect data wherever it lives – in the datacenter, at the edge or in the cloud. He highlights fast-start use cases such as customer-support chatbots, docum... Read more

explore Keep Exploring

What was the topic of the keynote interview at the RAISE Summit 2025? add

What is the role of enterprises in consuming and building applications using models, and how is Nutanix addressing their needs? add

What are the components of the relationship with NVIDIA and how do they contribute to the enterprise AI stack? add

bolt Powered by CUBE AI

Rajiv Ramaswami, Nutanix

search

>> Welcome back everyone to theCube's live coverage here in Paris, France. I'm John Furrier, your host of theCube. We're here for two days of wall-to-wall coverage of RAISE Summit 2025. We've got a distinguished Cube alumni on. Rajiv Ramaswamy, CEO of Nutanix. Rajiv, I feel like we just saw each other a few months ago at Nutanix next in D.C.-

>> Indeed. - ... but we're in Paris, France.

>> Welcome back to theCube.

>> Glad to be here. - Good to see you.

>> In a great city, Paris. - Great city.

>> You just had a keynote interviewed by a leader at NVIDIA

>> on the main stage. Thanks for coming into the post- game review keynote. What was the topic?

>> Yeah, the topic was around enterprise AI. How can we make it easy for companies to go build all these generative applications, consume these models and run them? And what we talked about is how Nutanix combined with NVIDIA is enabling that full stack, the infrastructure stack as well as all the AI stack on top of it to provide inference endpoints for companies to build these applications.

>> NVIDIA's got their hands in a lot of action. Obviously they've been dominating on the supply chain. Dave and I were just reporting last week, Intel's market share dropped from 73% in X-86. Now it's like 13, 15%, it's in the teens basically in one year, which is incredibly that that could happen. It just goes to show that the GPU side of the market is booming. This event here in Paris is speaking to entrepreneurs, developers, builders, neoclouds, infrastructure leaders like you and Nutanix where you have that enablement on that supply chain enabling the AI infrastructure. And then the next layers of the stack are tooling up and scaling up, and then that's enabling in a wave of applications and builders. You're on that layer of the infrastructure layer so great partnership. What does it mean to customers? Because at the end of the day, do I have to do anything differently? Let's just add it on top of Nutanix. I love Nutanix. What do I do with it? I want to go to Nutanix. So as customers make these generational decisions-

>> Indeed. - ...

>> because we are talking about right now, large enterprises, large cloud companies are making generational calls on the system-

>> Indeed. - ...

>> that they're building so it's a one-way door, but they're doing their homework. What's your answer to that because this is a big deal?

>> The vast majority of enterprises, customers are going

>> to be consuming models and building inferencing applications. They're not in the business of creating these models or training these models. That's a handful of companies that have the capital and investment to do that, and we know who they are. But here we have this broad universe and that's really where value actually gets created. If you think about it, this is where real applications get built that can change the nature of how you run your business. And so the vast majority of them don't really have the wherewithal or the sophistication to really build all of this stuff. They want a turnkey solution. And so that's what we are doing in Nutanix. We started out by making the infrastructure stack really simple. Everything is hidden under the covers. We hide a lot of the underlying complexity, make it really simple for companies. We used to call it invisible infrastructure back in the day, but it really is about simplicity. And then now the new layer is an AI stack that sits on top of it to create these inference endpoints. And that's just where we partnered together with NVIDIA and others to say, "Okay, now we have an infrastructure, platform on top of it we have a Kubernetes platform, and then we have the AI stack to really bring it all together into a turnkey solution," where a company now who wants to build an AI application can simply say, "I got everything else covered. Here's a shared infrastructure that I can use just like I used to run my regular applications, I can now program to an API that this provides and build my applications and run them."

>> And that's an innovation strategy that flips the

>> IT blocking and tackling to a value creation extraction formula.

>> Yeah, indeed. And by the way, we've seen this over multiple generations. Whenever things start getting mature, IT needs to be able to run it and operate it at scale to be able to deliver value to that business. And that's really what this is about. We are in the early stages of enterprise AI adoption and it's getting to a point where now again, and IT teams have to embrace this, figure out how to stand up this infrastructure and be able to deliver shared services to their internal clients that can then build these applications easily.

>> You mentioned the rich, my words, you didn't say rich,

>> the rich companies, the banks, and we know who they are, JPMorgan Chase had a 17 billion IT budget. Not a lot of other people have that, so let's call them the one percenters, the enterprise, but the rest of the enterprise, they're doing cool things. So they have IT. They bought over the years, many generations of servers and storage fabrics, storage systems, probably network attached storage. They've always been foreclosed and left out of the mix. Yeah, they've probably got some homegrown apps. They've probably got some Salesforce here. They got some off the shelf applications. They've been in this lull. This is my research, our research we're pointing to where, yeah, there's some examples. They do some cloud native stuff, but not actually writing transformative apps. I think that's what you're referring to is there's a sweet spot-

>> Correct. - ...

>> of market that's untapped. Is that what you're referring to?

>> Well, I think first of all, we include the 1% as well. So the big companies, the most sophisticated ones,

>> which have huge investments, big software development teams, of course, they're going to be the ones that are going to be building thousands of these types of applications themselves. Now, the others can still actually very quickly make use of simple use cases. So one of the most simple use cases is around customer support, and it's very easy on a relatively small infrastructure, a four node cluster to be able to build a customer support use case, a chatbot using standard open source models and put these things together. It doesn't require a lot of effort. So the simple use cases, document summarization, translation, chatbots, those I think can be easily put together. And even content creation, you're going to have third- party people, third-party providers for example, for software development, you've got a bunch of companies out there. So there'll be pre-packaged providers as well for AI software that companies can use also. Now, what they need though is to be able to run these wherever their data is located. And the data can be anywhere and they need to run it securely. They need to make sure that their IP is protected. And so some of this is going to be run in the data center, some of it is going to be run in the edges, others will be run in the public cloud. And so they just want to be able to consume and build these apps and run them simply. So there's a huge sweet spot for broad-scale adoption.

>> All right, so I didn't mean the pigeonhole Nutanix is

>> going after the underserved. What I was trying to get at, and again, a good call-out on that, yeah, you're a public company, you got to make your number, you're doing well, you got those companies already. But I think it's a growth strategy. Would you say that that's a growth opportunity for Nutanix as a business?

>> Indeed it is. - So it's not necessarily,

>> you're not doing any pivots.

>> You already have a great customer base there.

>> So today we have 28,000 customers, that includes a thousand of the top 2000 globals in the world. So we have the big customers. We have the small customers. We are across every vertical. And all these customers are going through a journey. They all have these bulk of their applications today, virtual machine applications. They are slowly but surely moving into a world where more of their modern applications are containerized, and Kubernetes is a platform for doing that, which we provide. The next phase in this journey is almost all the applications they run are going to have some form of AI embedded. That's the next 5 to 10 years, and that's just at the beginning. So for us, it's just about enabling all the applications that our customers want to run on our platform.

>> And I think that's the point, the take away is that the Gen. AI apps sit on top of what you do. You make it easier for them. You mentioned NVIDIA. I know , you have a really strategic relationship with AWS.

>> Indeed. - I talked to Tark and Maynard about

>> that the other day.

>> That's going to stream live tomorrow on theCUBE, our AWS mid-half-time report on the cloud. You got cloud, multi-cloud or distributed computing hybrid cloud, you got all the customers, you got the Gen. AI wave. What's the NVIDIA piece? What are you working with them on? Is that enterprise go-to-market? Is that enterprise functionality? What specifically around NVIDIA?

>> So there's three components of our relationship

>> with NVIDIA, and we've had a long history with them dating back to days before AI where we were virtualizing GPUs for virtual desktops. But today the first component is NVIDIA has what they call NVAI NVIDIA AI enterprise stack. That also includes their microservices engines for inferencing. So we include all of that in our platform in the AI stack that we can provide to the enterprise, and customers can put that in, put those together, have automated workflows to download models from the repository and do that. Second, NVIDIA has a reference architecture for compute that's based on top of Kubernetes. We are part of that reference architecture. So they have a similar thing for storage. It's called AIDP I believe data protection storage, so we are part of that as well. That includes all the storage elements, vector databases, et cetera. So we are a fully validated design partner of NVIDIA, and we take elements of their stack and customers can use those elements on top of our infrastructure to get a turnkey stack.

>> So you're basically taking platform engineering concepts and infrastructure concepts that you guys have been doing well in, bringing NVIDIA into the table to bring in all the AI goodness that they're coming in with so that those enterprise can scale up.

>> Exactly. Think about it as building on top of our Kubernetes stack. We already have a Kubernetes stack. Now all of these AI stack elements are being built on top of that, the different kinds of models that people can download now, the security and governance elements of that that we provide, the automation of the workflows. So that's the value that we add on top of this to be able to provide that turnkey solution for our enterprises.

>> What's interesting about what you guys are doing, I think it's the first of all, great strategy. I have to give you my thumbs up on it, but because you have done the hard work, it's still not going to go away. It's just going to be abstracted away. But now Gen. AI is going to enable the agent layer, which is going to create more-

>> That's right. - ... services.

>> That's right. - So you've got the data checked off,

>> you've got the Kubernetes, which is cloud native,

>> that scale out and scale up with the Gen.

>> AI. As you look at customers, what are they doing? Can you share some examples of where they're innovating and where they're extracting the value? I love the Nutanix stack AI stack because in the enterprise, that's been a real problem. Kubernetes has actually done great.

>> Yes. - It's de facto rallied

>> and become an orchestration layer for cloud native services. But the enterprise has been holding like there's a blockade, there's a blockage of POCs trying to get into production.

>> That's exactly right. - This is a

>> problem that's getting solved. Sounds like you're solving it.

>> It is. We've seen some simple use

>> cases go into production. I will put them into three categories. The first category is just simply around summarization of content, summarization of content, automated translation of languages, et cetera, that's a very simple category. That's something that simple LLMs can do really well, first use case. Second use case is around customer support and chatbots, another simple use case. By the way, these can be done with very small clusters. It doesn't require massive investment in farms, small clusters. That's the second use case. The third use case is content creation, whether it be creating code, whether it be creating documents, whether it be creating video content, whatever it may be. These are three relatively simple use cases for enterprise AI. Now, in all these cases, the data may be on-prem, the data may be in secure locations. You need to provide governance. You need to provide security, roles-based access, control, and protect the IP so those are all critical things that you need to simplify it. So these are three simple use cases today. Now, this stuff is becoming more sophisticated though. You can see a world where you're going to have much more sophisticated multi-agent workflows that get enabled on top of a platform like this, and that's coming, not quite here yet, but it's coming.

>> I have to ask you because I love interviewing startups because they like to throw out the haymakers. I interviewed a startup earlier , I won't say their name. People could go look at the videotape, video disc or stream. He said in three years, everything that we do is going to be automated. Now you live in the automation world. So I have to ask you the reality, he's mainly talking about tasks with agents, but it's been hard to automate end-to-end workflows to get it 100% accurate, because you really don't want to failure. Forget the hallucinations that's more on the top of the stack, but take me through your mindset on how you see automation accelerating or not accelerating. What's your view on, you can comment on the three- year mark if you believe that or not, but he said three. That's a startup. What's your view on automation?

>> First of all, it's a continuous process. It never ends. It never ends. So this notion that we are all going to be done in three years, no. It'll continuously be going on. I'll just give you an example. I'm picking it off the cuff. As a software development company here, yeah, whenever we develop new software, of course we write test cases for it. The first time you do it, you have to write it yourself, but now we can potentially use AI to start generating those test cases. But after that it's automated. But you're going to create new content all the time, and some of that stuff may be manual, but over time, more and more will get automated. So it's a never-ending thing on that front. But I'd say one more thing. It's not just about automating existing workflows. It's about how you think about reimagining the workflow itself because the workflows don't have to be the same. I'll give you an example. This is not original. I heard this from a McKinsey guy one time ago. Think about it, if you're interacting, for example, between three different people, maybe there's a designer to mark it up, and then afterwards writes a set of requirements, and that process could take weeks to months. Now all of that, you can actually specify what you want from a workstream perspective, from a flow perspective, all of that can be auto- generated and done very quickly. So you don't need these three people doing these things. You can actually do them all in a much simpler way. So that's just one example of how a workflow could be reimagined.

>> So automation really is about efficiency of task.

>> Absolutely, yes.

>> So that's what you're getting. It's not so much... We're

>> like robots.

>> And reimagining how these tasks have to change. You don't

>> have to do them the same way that you're doing before.

>> Rajiv, it's always great to have you on theCUBE and congratulations on a great keynote and your business success at Nutanix. You guys have got a lot of customers. I have to ask you a personal question, and I've interviewed you many, many times. You've seen many ways of innovation.

>> Yes. - You have a great technical pedigree.

>> You've written books on the topics that we've talked in the past. Given all your internal knowledge and instincts, scope the opportunity that's in front of us because we're seeing such a major shift. You could look at any dimension of it, the data side, the automation, the ages, how that's going to have to be delegated and deployed, the operating system side of it. There's so much tech and computer science now. There's so much business transformation, again, business transformation, technology transformation, societal transformation. How do you see the future, I guess from your personal perspective, because if we were in college again and we said, "Hey, Rajiv, let's do a startup. " What would we do? We be like, it depends. There's so much to go after.

>> Look, that's exciting about tech. I've been in tech now for 35 plus years, and it's never... still, it's always evolving. There's always a new thing that comes in, and these things take time to shape. There'll be ideas and then there'll be investments. There'll be some wrong starts. There'll be some losers and winners. Then eventually winners, and then once it's established, then it starts taking root over a period of time. We saw this with the internet a long time ago and in 20 plus years ago, and there was a big boom and there was a bust, and then there was a steady uptick, and now it's such, we don't even think twice about it.

>> It all happened. - It all happened.

>> Now it's everywhere and we don't even think twice about it.

>> It's the same thing with AI I think we are in that same phase now where it's huge potential, I think long- term to change how we think about things in every way, but we past the early stages now and it's evolving at a very rapid clip. We talked about LLMs and we talked about inferencing, and now we are talking about agents and this thing changes every year, and that's exciting for a tech guy like me to keep up with it first for all and learn about it, and then figure out how to eventually make this stuff be available in production. As an industry, we have to figure out, it's not enough to think about cool ideas. It's about how do we actually enable it and enable people to become users of this stuff.

>> What jumps out at you technically? I know you've done a lot of tech, you're a deep techie at heart. Are you coding still? Are you vibe coding?

>> It's been a long time since I've coded. I don't think I

>> should ever code at this point. In fact, frankly, most people are not going to do actual coding. They're going to figure out how to use these models and do .

>> I actually vibe coded two weekends ago, didn't write a single line of code, and it was actually fun. It was entertaining. And the app, I built something I really wanted to build, but I actually did the low level-

>> I do. I play around with ChatGPT or other tools all the time and tell it what I want to do and that it's not bad.

>> As a techie, what's jumping out at you? I know you have a deep tech background. Are there any things that, it doesn't have to be Nutanix plate, is there anything that's popping up that really gets your attention that, "Man, if I had free time, I would dive down that rabbit hole. I would really explore that technology. " Is there anything you see that gets you personally energized?

>> I would just say I think the potential to see broad-scale adoption. It's not one piece of technology. To me it's actually about how do all these different pieces of technologies come together to create something that people can use? Because you see a lot of concepts being thrown around, but how do people actually get to use them and simplify them? It's more about that for me at this point, and that's really what Nutanix's philosophy is about, trying to make these things simple, hide the underlying complexity and make it useful for people.

>> I can tell you're an operating systems guy. That's what an operating systems thinker thinks about. Rajiv, great to have you on theCUBE.

>> Thank you, John. - Thanks for closing out our day one.

>> Rajiv here, the CEO of Nutanix CUBE, friend, alumni,

>> and deep tech leader closing out day one of theCUBE here. Our stream is wrapped up. We'll see you tomorrow on Thecube. net. I'm John Furrier, host of theCUBE. Thanks for watching.