 from Washington D.C., it's theCUBE, covering .NEXT Conference, brought to you by Nutanix. Welcome back to Nutanix.conf, everybody. Sorry, .NEXTConf, hashtag NextConf. This is theCUBE, the leader in live tech coverage. My name is Dave Vellante, and I'm with Stu Miniman. We go out to the events, we extract the signal from noise, we have a real treat for you. The day two keynote was done by Sunil Poti, who's the head of product, chief product officer in development at Nutanix. Long time, CUBE guests. Sunil, good to see you again. Thanks for coming on. Thanks, Stu. Good to be here. I agree with Stu. Great energy in the keynote this morning. I have to say, we got in there, Stu, at around 10 of, the place was not packed, but by the time you started, the place was packed. So we heard the Lenovo party went to like 2 a.m. Got it. There was some excitement. There was some excitement. We came in and yeah, a lot of cheering. So you must feel pretty good about that. Yeah, yeah. I think, as I was telling some of you guys, I think right now we are in our growth years where it sort of feels like we've got a lot of fans because folks want to sort of relate to technology, companies that foundationally are disruptive and keep disruptive, right? And it helps folks' careers, it helps their personal lives, everything, right? So the vibe that we get out of these dotnecks conferences are mostly about this. They're not so young people, they feel young using the technology at work and while at the conference. So that's, I guess, part and parcel of some of the, sort of the excitement that you're seeing probably. Well, so let's get into it. I mean, a lot of announcements this week. Where do you want to start? So I think, that's a big question. There's a lot of stuff. Well, maybe we start with the strategy, which we laid out two years ago. We're not just HCI, we're going cloud. And so that's sort of the setup. And you had a number of proof points and product announcements this week that underscore that strategy. So I think I'll break this down into maybe four parts just to structure it, which is obviously we've moved maybe on HCI, we've subsumed virtualization and then we set cloud, if you remember, 2015. And then we improved on it in 2016 and so forth. The big thing was, you know, every few years it gives you an opportunity to truly offer a step function change in transformation of the company. And that's what 2017 for us is. That's why the announcements are also packed. It's not just an incremental set of things. So let me break it down into four segments. One, the first thing is the fact that it truly is clear to us and through our customers that this needs to be this single, software-centric fabric that gives you optionality of any hardware platform, any hypervisor, any consumption model, whether it be a pay-as-you-go, you know, it could be appliance driven, it could be pure software, real aids and so forth. And then obviously with our partnerships with, you know, Dell, Lenovo, support of Amazon and Azure and now Google in a big way from a public cloud side that I'll come to. But then extending that to our software support on Cisco and HP and now with IBM power especially, changing the game in terms of our enterprise workload. So the first sort of segment of capabilities is to really make it look like our platform story is becoming an OS that cuts across, in fact we call it all of our applications, deployment form factors while being open. So that's the first category. Let me just maybe quickly summarize it and then we'll come back in a minute. The second is the fact that look, while we were doing that, we were still taking an infrastructure-centric view, right? Whereas VMs, containers, whatever it is. So now what we've found is that look, you know, fundamentally we need to change the operational construct, elevate it to be an application-centric world and that's where it comes in. Independent of hybrid clouds, just from a private cloud side itself, just even if I'm just elevating my current infrastructure to a private cloud, an app-first automation is a good thing and then what we've done is in that second category is to merge quote-unquote on-prem infrastructure with off-prem using this multi-cloud thing. That's the second thing, that's good. The third thing is the fact that, well, guess what? While we can bring provisioning and operational convergence with calm, true lift and shift is still very hard because the stacks on both sides are different between public cloud and private cloud, right? And that's where Xi comes in, which is our new cloud services. Essentially it replicates your on-prem stack and seamlessly extends your data center so you can do some things like one-click DR and so forth. And the last but not the least was as we were building Xi and we get into it, Google and Nutanix have started getting really close from a technology integration and a delivery perspective where things like Xi could become ubiquitously delivered, but more importantly I could take Google Cloud past services and fuse them with Nutanix enterprises. So those are the four. So Sunil, let me poke at something. So we've heard the story before, there are lots of choice to build an ecosystem, but AHV seems to be a strong component, so Xi, you got to use AHV, got to use Nutanix replication. It's like, you got lots of choice unless you want to use all these cool things we're doing, in which case you want the full Nutanix stack. So I think that's a constant thing for us is, look, at the end of the day, to us lock-in should be by choice, not by need, right? So it's up to us to offer a series of value-added services on our full stack that customers choose to go to that full stack for value. It's just like on-prem, right? I mean, look, 60, 70% of our enterprise workloads are still ESX-based. Hyper-V is still there. Under the cover, Supermicro or NX is no longer 100%, it's coming down. So frankly from a customer perspective of deployment, that's where they start. Maybe they'll stay there. We can need to add value to the environment, but if you use AHV, you get microsegmentation for free. It's one click. It's up to the customer to choose to use AHV based on the choice at that point. The same thing applies to Xi as well, saying look, on day one it'll replicate on-premise to off-premise, but on off-premise, at least in the enterprise side, it's to, we will still support ESX and multi, you know, essentially the open environment is still a source for us. But target, on day one at least, we can build a honest product without being completely integrated. Yeah, and Sunil, definitely there's choice there. Even look at the Google announcement, well, we've got the AHV solution tour, if you want to go down the containerization, Kubernetes, that that's going to be somewhere. Maybe, you know, we've been positing that the partnership with Google should help you accelerate that move towards containerization. I know it's early for a lot of customers, but any commentary you can give on virtual search? In fact, I would say containerization, you know, sort of expanded to the broader cloud native workloads aspects, which is to us the big thing that customers came to us, and we've seen that now resonate is, they just don't want DR, obviously, one click DR is a big deal, if you can pull it off by replicating the stacks. But really what they want from the enterprise side is, they want to take enterprise apps that they are elastic, consume them as a cloud service, but co-locate them in such a way, not just in the data center level, but at an application operations level, application integration level, so that they can core recite as if they were on the same node with a set of Google pass services. So essentially think about it, I've done all this work to do one click DR, I've moved warehouse management system, running SQL server or something else into the Nutanix cloud, that's a little bit easy now with XI, but now I can now run a BigQuery app, I can run a series of services as if they were in the same VLAN, right? And that's the real power of like, I love that and I guess the easy compare on that is later this year we expect VMware and AWS to run. I'm sure you would posit that, VMware and Amazon pricing might be a little bit different than the Nutanix and Google offering and what services and how you get a little different. Yeah, I mean, I think I'm sure you get a lot of responses on that one, but I tell you my take is actually, even before comparing the approach, so first of all, just philosophically, I think the strategy makes sense of trying to make hybrid invisible in general, right? I mean, VCloud frankly was the first cut at that. The way we look at it was, VCloud was the right use case, wrong implementation. And to me, I think that is still the most important thing, which is before we build this hybrid cloud, whether it is with AWS or Google or anybody else, we'll talk through it all with Nutanix or VMware, you still have to build a proper private cloud. As in the cloud that you powers your primary data that needs to look like a true Google or AWS. And to us, unless someone re-engineers that stack, that's going to not be the same as, oh, even if I took a, you know, if I can call it a half-baked private cloud and extended it as a service, it's still half-baked hybrid cloud. So that's going to be the big thing, I think that's going to turn things out. And that led us at Wikibon to coin this term, true private cloud, we get a lot of grief for that term sometime, but we saw a lot of cloud washing and the concept is basically to substantially mimic what's in the public cloud and then create a control plane that spans multiple physical locations. Now, one of the things I've been getting a little grief on in this show, because we think Nutanix is an instantiation of what we call true private cloud. So one of the folks in our community was been hitting on me and he actually wrote a piece, this guy you're on, Javier, he's a very sharp guy, said that guys like you and others have no chance against the public cloud because he said this, HCI is stateful, I want to read it and get your feedback. HCI is stateful, it's VMs, it's V disks, they're like IT pets with a lot of labor. AWS is stateless, with microservices, it's object, it's database as a service, AI as a service and it's built for devs. I know you understand this, but how do you respond? No, I think look, frankly, we had a choice as a company a couple of years ago, when we were growing, the primary question for this company was, well if the apps are moving outside, forget about IAS or AWS, it's SaaS and the users are moving outside, again forget about dev and object store and all that, it's about consumerization because of mobility, everybody's got a phone, they can access an app, who worries about infrastructure, right? So as an enterprise infrastructure company, that's a secular question to answer, right? So for us, there's a reason why we didn't pack up and sell the company and move on, okay? So there's a theses for the company, the theses for the company is that we fundamentally think that cloud is not a zero sum game. And we're seeing that not just with the largest assets, like all the big dot coms that went with cloud, defined cloud and they're coming back now, right? One of your popular, whether it's your, you can call it your consumer devices that actually started with their service completely, as they grow, they're actually moving half of their services back to the private cloud, right? I think but it applies to the mainstream enterprise, which I call the fact that because public cloud is not a zero sum game because not because of security or compliance because those are temporal things in my opinion. The moment AWS puts a data center right beside my data center, security is a little bit in our compliance or data regulation is a little bit avoided, right? The real reason is purely going to be a financial choice for the kind of workload. If it's a predictable workload, even if I could re-engineer it, if it's predictable, it's kind of like me coming in, staying here in DC for three days, I'll rent a hotel. If it's there for a year, I'll lease an apartment. If I'm there for five years, financially, Excel makes certain math real, right? Doesn't matter if my costs are cheap. It's going to just work that way. I can buy hardware, capexize it, amortize and so forth. So I guess the simple answer to it is eventually I think there'll be 100% of the market will move. In some markets it'll be 70, 30, in some markets it'll be 37. We are at 1% of that market. So for Nutanix, the more someone tastes the wine of public cloud, the faster they'll accelerate the transition of their private infrastructure to look like Amazon. And in that sense, we are like, look, that's why we want to do XI, is because XI actually takes DR infrastructure away from us, right? We're probably making more money selling Nutanix in the DR data center, but we are accelerating the move because we think that that eventually accelerates every workload to come to the primary infrastructure on Nutanix. So at the end of the day it's going to be not about objects versus VDIS, I mean, he's right also in the sense that shame on us if we don't have a level of abstraction that is app-centric, that you don't have to really care about whether it's an object storage or a VDIS or anybody, so that whether it's a developer or an IT operator, they use the same operational interface. All right, so Sunil, we like that Nutanix is putting out a vision, my understanding, you know, the cloud service like next year, it's going to be shipping. What I'm a little worried about is we're almost out of time with you and you went through so many different pieces. You know, there, of course, there's the calm, we're going to talk to a detail a little bit, but maybe give some of the highlights as to the stuff that shipping now or soon that your customers are talking about. So what's happening is basically we take those four segments, right? The core software fabric evolving across every platform and so forth, there's a bunch of stuff that has already started shipping in the 5.1 release which came out a few months ago and a lot of it will ship in the 5.5 release that's coming later in the year. Calm is part of that release, as you guys know, it's part of the same, it's baked in. We had a choice, basically Calm has been engineered a couple of times over six, seven years. It's not a 1.0 product, but we took the time from last year, we didn't release it, created a Frankenstein, a power-sucking alien on the side, like some other tools. We took the time, to be honest, to integrate it into the console, you saw it, I mean it's absolutely taken the time. Functionally it's there, but that'll be part of the same release and then the XI project will be early 2018. The Google integrations will come in a stage today, even as early as later this year with Calm and Kubernetes and so forth and then extending to XI. So the time frame that we're talking about is probably minus three months to plus nine months. So the DR solution that you showed though, that is native Nutanix tech, right? That's not partners. It's Nutanix software delivered as a full stack and this is another thing that we had to take a hard call, it was a raging debate a couple of years ago, it was like, who builds a public cloud in these days? And that'll be the obvious question and the question was really, it's not about building a public cloud, are you building a cloud service? Whether it's on-prem or off-prem, are you being honest in the product design, right? Do you do billing and metering in one click? We don't do that today, but as building a cloud service through XI, we're building all those for our private cloud customers too. And so the goal was, kind of like six years ago, the easy answer was to take Nutanix, sell it as software and save your VMware for the new era, not worry about BIOS firmware, upgrades and all that stuff. And so until we did Nutanix on Supermicro, people actually believed that the market existed and then Dell and Lenovo and others came to market. So same thing applies with XI in an accelerated way. The moment we had the conviction to go build a full stack, make it look like an exact replica, but build these cloud capabilities. That's the big thing that I think some of our competitors didn't do was they took the same stack that wasn't a cloud and quickly tried to make it a cloud service, right? So that's the reason why we are starting with our own service and data centers and then scaling through Google, for example, right? As a way for us to not just get global reach, but also to merge cloud native apps with these enterprise. But the other obvious question, and I'm sure you guys had a conversation about this internally is some of the folks in our ecosystem are in that space, you know, is this competitive to what they're doing? You guys have always been a customer problem solving company, but what was the conversation like in that regard? So it hasn't come up to be honest. I think the cloud service aspect is still relatively new. It's siloed, I mean, we're pretty clear, right? This is not an end all be all service. We don't want to overreach from our things. It's about, look, you first moved to Nutanix on premise and it's your choice to say it should be as simple as my provision VMs and the more VMs I have or containers on Nutanix, I should right click and say protect. And just like iCloud for the iPhone, they're in the cloud, right? That's the goal, the true product goal. But Amazon like experience that you're trying to get. And it's a service that's integrated into it. And so when we have partners who have been hosting Nutanix, if you're talking about service, we've already had questions on like, rather than just create yet another service, provide a partner program that automatically conflicts and all, you know, eventually what we think is that we will offer turnkey services, not a generic purpose infrastructure. There's always use cases where somebody wants to outsource their primary data center and that's where a lot of our partners look. Especially the mid tier partners are in the mode of outsourcing the full data center. And for them, offering one click DR on their primary outsourcing is another advantage. He's too good, we're not going to let him go yet. So yeah, yeah, yeah. That's a good one, yeah. So Prism Central, some of the other announcements outside of the cloud pieces, maybe just touch on some of the highlights. So I think they're all part of the core fabric and fortunately they get, you know, subsumed into the, you know, when you have big announcement, they kind of subsume. But a couple of interesting that came up is, I think, and this is particularly, I'm pretty fond of a couple of features. One is Acropolis file services. You know, I made this joke about killing NetApp a couple of years ago. People keep reminding me about it. I think it's one of our fastest growing features in the last six to nine months. And it's a natural sale for us. People go in there, it's been a little bit limited because of SMB functionality, but now with NFS. It opens up that. Another one, though, that's more, probably more secular, is this concept of machine learning now coming into mainstream. People talk about AI and all that. And it is a lot of churn for us, but it all comes down to manifesting itself as tangible things that the customer sees. So for example, if I can show that certain amount of downtime genuinely reduced from four nines, it became five nines, without the operator having to do anything, then it is becoming intelligent. So I think our bar for machine learning, which you will see more and more, by the way, as a secular thing, I think I'll predict that, not just from Nutanix, AI, if I have to use the word, is going to become more prominent in the next few years, just like cloud was. So Sydney, the last question I have for you, from a development standpoint, where won't Nutanix go? Where will we not go? Good question. So for example, there's an obvious thing like AWS says, oh, let me instrument every service that sits on my platform. If I like that service, I might end up doing it, right? The classic, if I call it the tyranny of being a platform partner. See, for us, I think because of two things. One, our ceiling of our current market is super high right now, Stu, right? I mean, we still, yeah, it's a billion bucks, growing 50, 60%, whatever, but we still, even if half the business goes to AWS, Google, and Amazon, we can still be a company larger than we ever are. It's nearly a trillion dollar market. Yeah, I mean, it's a big market. So therefore, we don't have to be greedy about near-term. We can be long-term greedy. All right, we got to go. All right, thanks so much for coming on. Thanks a lot, guys. I really appreciate it. All right, keep it right there, everybody. We'll be right back with our next guest, right after this short break.