 Live from Miami Beach, Florida, extracting the signal from the noise, it's theCUBE. Covering .NEXT Conference, brought to you by Nutanix. Now your host, Dave Vellante and Stu Miniman. Welcome back to Steamy Miami everybody. We're at the Nutanix Next Conference, Stu Miniman and Dave Vellante. Ravi Matre is here, he's the founder of Lightspeed and Ventures, awesome VC. You've got some really terrific investment app dynamics. Obviously, Nutanix, Rubrik, we were just talking about them. Cubilum, a premium. Really doing some interesting things in security and many, many others. Ravi, welcome to theCUBE, thanks for coming on. Yeah, thank you for having me guys. Yeah, so keynote this morning, up with Vinod and impressive thousand people almost at a inaugural conference. We've not seen that before. We do a lot of these events. They tend to be little hackathons and small little meetups, but it's a big deal. Yeah, yeah it is, I was as impressed with the size and the energy in the audience. We work with a lot of startup companies. I'd say this was probably the most well attended inaugural user conference event for a cross portfolio of 200 companies. So yeah. So the risk of being a little bit repetitive of the keynotes this morning weren't broadcast live, but what intrigued you about Nutanix when you first saw their founders? I mean, describe what the first meeting was like. Yeah, we obviously work with lots of companies in light speed. Our entry point typically is at the earliest stage when it's usually a handful of people or maybe even one person with a big idea, but it's at the very formative stages and so we're really looking for, a lot of it is about trying to evaluate the people and understand the boldness of their idea, the possibility with our support for those small set of people to really go execute on the idea and make something big. And I would say with Nutanix, and I said it this morning, one of the things that really impressed us about the founding team was not just the boldness of their idea in everything that permeated, there's lots of companies that talk to us about storage, about doing new things in the data center, but it's this notion that taking complex ideas and making them simple is really, that's the sign of true genius. And with Nutanix, their vision from day one was to take extremely complex technology across the stack and bring that together in a way that was exceedingly simple for users. We talked about it as a partnership and we said, that idea if they really can deliver on it and build a product that embodies that core principle, it will be disruptive to computing in the data center because historically there have been lots of technology innovations, but with respect to a company that really understands how to make those innovations easily accessible to users, there have been very few. I mean, Apple is probably the best example on more of the consumer side of utilization of technology, but in the core of the data center for systems admins and database administrators, people who do the hard work, it's not very sexy. There hasn't been a company in the recent past that's figured out how to make those technologies really simple and easy, so when we met Nutanix and we met the founders five years ago, that vision really, it captivated us, it captivated our imaginations. Yeah, so when you look at, I was thinking, because Vinod said we haven't seen a major architectural innovation since Sun, and I'm thinking about what you're saying here and I'm thinking to myself, well, what about VMware? But really what VMware did is they took something that was incredibly complicated and inefficient and made it more efficient, but it's still really complicated, and you're saying Nutanix is taking it to the next level. Yeah, they made it more efficient, but it's still a fairly complex set of tools and technologies for the mere mortal to try to turn the knobs on. And really as we go into the future, I think Deraj talked about it, the applications are changing, they're becoming even more complex, there's more pieces to an application, you hear about containers, you hear about microservices, that just means there are more units of computing workload that a single person has to try to control and manage. And so with that increasing proliferation or sprawl, you need, the challenge, the bar from creating simplicity, it gets even higher, because you have that many more things that one or a few number of administrators needs to be able to manage without having to think about it. Yeah, Stu Moritz even called it a software mainframe. So I mean, there you go, that says it all. I mean, we watched for so many years, there was kind of the disaggregation of pulling all the pieces starts, everything became there, cylinders are excellent, or bespoke silo of pieces together. And they said here, it's not about convergence, but it's about totally re-platforming as opposed to just putting in a layer of abstraction. Yeah, I think that's right. And I think it's interesting, the company talks about this notion of invisibility. It took a while for that to really grow on me, but I think it's a pretty appropriate analogy in our lives, things that become invisible, that doesn't mean that we don't value them or don't appreciate them, it's just that they make our lives sort of simple enough that we can move on to the next thing, and tower-dusing analogy of payments, Uber's made it, so you don't even think about paying for a cab anymore, it just happens in the background. Yeah, so one of the biggest challenges we have in IT today is we've got kind of that legacy, we've got those old applications and making that migration is tough. You know, I think one of the theses is coming out of this conference is if we can change that underlying substrate and make it so that it's ready for both the old and the new, it should be able to unchain us from the ball and chain that was infrastructure. Yeah, and I think one of the things where Nutanix has had a lot of foresight, I mean, the world we live in today, in terms of computing and the enterprise, there's absolutely a movement under foot for computing to move to the cloud, to the public cloud. And you have some major, major companies in Amazon, in Microsoft, in Google, who I think have built platforms that rightfully can support some amount of enterprise workloads in the public cloud. However, the way in which those companies are advocating for enterprises to move workloads to the cloud, they've set up a little bit of a bright line. You can't, if you're going to move the workload to the cloud, you really don't have the ability to retain any elements of how you would manage or deploy that workload at all. You have to trust the third party to do it. And so enterprises are sort of stuck in this place where it's bimodal. They either completely abdicate responsibility and put the workload in someone else's hands, someone else's hands, or they retain it with all of the complexity that they have today using kind of technologies that are hard to manage. And Nutanix has, I think, figured out how to say those two worlds shouldn't exist with a bright line between them. Let us provide a vision of the future where you can use one or both of those in whatever proportion you need, but it takes a really new piece of technology to do that. Yeah, I think we're going to see real blurring of that line that you talked about, because if you listen to Amazon, they've been talking about, oh wait, no, you can pull your policies, you can pull some of your ownership pieces into what we're doing. So you really see those hyper scale guys trying to pull over to the enterprise just as you've got companies like Nutanix that are saying we're going to pull the enterprise to more of that web scale model. So absolutely, we've got a little bit of confusion and therefore we have this kind of hybrid offering which is where things like XCP and Acropolis can help customers because this whole migration is one of those things that, this is a five to 10 year move as to where is the equilibrium, what apps live where, and I don't want to be stuck in the past or over pivot to the future. Yeah, that's right. And I think the interesting thing for us is that most of the large income in vendors, they don't have the DNA or the wherewithal to reinvent their technology stacks to address that issue, which creates tremendous opportunity, obviously. So you used the term doom loop. I loved it, that was my favorite phrase of the morning. I was tweeting out to explain to people what you meant is this sort of incremental mindset of incremental investments. And so many examples. I want to ask you though, so some of the companies are getting pretty good even inside the doom loop. I take a company like EMC who's a target of Nutanix, clearly, but they do a good job, a great job of making incremental investments to their VNX and their symmetric lines, just bringing the customers along to the next generation and then grabbing companies and bringing in tuck-ins to do inorganic innovation. Are we going to see another company emerge? I mean, you saw the wave of three parts and isolans and data domains and compalence and on and on and on who got the IPO, which is great for you guys, for investors, but they weren't able to thrive as a public company. Do you think we'll see the emergence of another great infrastructure company? I think that's not only possible, I think it's very likely. I mean, you mentioned EMC and I think they're a company, obviously, that's had a terrific run as have many other companies like HP, Sun Back in the Day, which was mentioned earlier at the conference. You know, what happens is every 10 to 15 years we go through these major, we like to call them technology upgrade cycles, so there was a movement from mainframes and minis to client server in the late 90s and then in sort of the 2000 to 2005 period that really transitioned to the Web 1.0 architecture, the data center. And here we are 10 to 15 years later. The data center architecture is fundamentally changing again. It's going through a major technology upgrade cycle and that's because the way that applications behave and the way they need to be supported by the computing infrastructure is fundamentally changing. You have componentization, you have microservices and that means that many of the traditional vendors, the EMCs, the HPs, the IBM, the Oracles, they've designed the computing data center in a way that it supports the last generation of applications where you have maybe a web server tier, a Java monolithic application tier and a database. Today you've got things spread out all over the place, API driven services, you could have hundreds of components that actually make an application. So to support that at the data center computing layer and storage layer, you need a fundamentally different platform and the challenge with all of these legacy vendors is that to support the new applications, they essentially would have to build from the ground up. And all of these incremental tuck-in acquisitions or incremental innovations on the old data center computing platform paradigms means that you're still stuck with this, hey I've built an architecture that supports, you know, largely it's sort of a three tier application and so what we see happening is that that's what creates the opportunity for a major new vendor, a new Tanix or otherwise, to ascend and to become that next generation computing platform that supports today's application. Because it's so ugly, you're right, I mean EMC's got a lot of complex stovepipes. If it weren't such a radical change in the technology stack, you know, I think that would give more opportunity for the incumbent vendors to make the transitions. But we're just in a period where the change is happening so rapidly. I mean the mindset of the users of technologies, the admins, the CIOs, the people who actually are tasked with getting applications out quickly and having agility, their mindset is I can't actually do this with the same old pieces of technology that I use. I must consider the new vendors, I must consider the innovators. And that again is, you know, you have to get it to a period in time where it's sort of, we call it the innovators mindset where the users of technology sort of realize that they have to look to the new innovators and we're in one of those periods right now. Well there's certainly no debating that companies like Nutanix can do very well and we think, we think, we put Nutanix, we're looking at pre-IPO, we're looking at service now, Splunk, Tableau, it feels when you talk to customers like similar pre-IPO momentum, you know, Workdays, another one, maybe they're, you know, even, you know, different space, obviously. But your premise, Ravi, is that you will see a multi-billion dollar company emerge within your core enterprise space. It's not going to come from Amazon or Facebook or it's going to be your core sort of investment sphere. I think that will absolutely happen. I think Nutanix is extremely well positioned to, you know, be one of the companies that can be, you know, sort of the category winner in this new, we call it sort of the new enterprise computing platform race. And yeah, there are a lot of reasons why we're sort of at that unique window in time where I think a new set of category winners will emerge, yeah. So I'm wondering, you know, if you look in the market in general, there's kind of the, you know, what's just brand new and what's putting together a lot of pieces that were there. You know, people say, you know, Waze didn't create any new technology. They took a bunch of pieces and kind of built it together. You know, Uber and Airbnb are just reinventing the old model. If I look at Nutanix, I mean, things like Acropolis, it's built off of open-source KVM. They're using commodity components. Obviously there's some core IP that they have in there but there's, you know, no proprietary hardware. It's, you know, lots of options that they choose there. Is that kind of a new model that you look at from an investment standpoint as to pulling those options together? Well I think a lot of times the innovation in the IP is it goes back to the first principles of the company is in taking the incredibly complex and making it simple. I mean, if you look at Apple, their core technology innovations were not about, you know, somehow allowing, you know, video music to be able to be streamed to a phone. It was more putting together those technologies in a way for where for the first time they became so accessible to the user that it was just intuitive. You know, Steve Jobs once said when he was talking about, you know, the first iPhones and iPads. He said with respect, and iTunes, he said it really gives a user, he talked about how, you know, when he was young he had hundreds and hundreds of vinyl albums and he just, he never could play them because he had so many of them. He didn't know where to start and he said, these technologies finally will give a person a way to rediscover all the music that they have. And I think it was the power of the integration of technologies that already existed that allowed that, you know, kind of breakthrough experience to happen. That's really, I think, the analog with Nutanix I think that we think about. It's just there's a lot of hard technology required to make all these disparate pieces of technology come together in a way that for the user, they just really don't have to think about it anymore. Yeah, it's interesting to me, because some people say, you know, how many open source companies we're going to have and Nutanix is a good example of leveraging open source in its DNA. Yeah, I think open source is terrific. It really does democratize and bring down the overall cost of being able to have web scale computing. You know, the challenge is, I mean, you know, whether it's trying to bring together things like Hadoop and Spark and Kafka and Flume and you know, I could go on, there's probably hundreds of open source technologies that one needs to bring together to have a viable solution. So the magic of Nutanix is saying, hey, we will allow the user to avail themselves of a lot of these technologies. Now they're very powerful and we won't charge the tax for using a Cassandra, but we will bring those things together in a way that, you know, you get an abstracted, much the higher level of power in the hands of a user and administrator. That adds value. So open source is the main spring of combinatorial innovation in this case and many others, but the flip side of that is, you know, it creates complexity. But yes, that's right, it creates complexity, but open source is a business model. We'll see, right? I mean, we talk to Rob Bearden about this all the time and you know, we're watching. There are, you know, few and far between examples. You know, at the end of the day, you know, open source, again, I do think it democratizes technology. When you get to, you know, production environments, particularly when you get outside of Silicon Valley where there just, there isn't the where with all the ability to hire, you know, these sort of extremely knowledgeable, you know, technologists who understand all the latest and greatest open source technologies and how to make them really work. So when you get to production environments for the typical mid or large enterprise, there has to be some vendor or set of vendors that basically brings all these pieces, orchestrates and brings these pieces of technologies together in a way where the user doesn't have to worry about what's going on under the hood. There's no other way for mainstream America or for the world that matter to digest these technologies. So Ravi, you and other VCs go to Amazon re-invent, you're looking for sort of that, you're watching that ecosystem emerge. And it's interesting to see Newtonics both embrace this sort of what we call inter-clouding, you know, multi-cloud sort of environment. But at the same time, it's down the road, you can see a collision course with, whether it's VMware or Amazon. What are your thoughts on what's going on in the Amazon ecosystem and what it means specifically to Newtonics? Well, I think Amazon is a absolute force to be reckoned with and they had the vision long before others about the notion that if you could provide a functional full computing stack where all of the underlying complexity of provisioning and managing it is abstracted from the users, that there could be tremendous value in that. And they've continued to add new layers of capability on top of the core virtual machine computing component. You've got things like Redshift, you've got EMR, you've got EBS. So a lot of the layers that let users now, again, think about higher levels of abstraction, they've moved in that direction. So I think the public cloud and enterprises using that as a place to viably run many of their workloads, I think that's going to be an important part of the market going forward. The reality is we're not going to get away from the fact that there are many application workloads that were built five years ago, 10 years ago, 15 years ago and these workloads will continue to run for a very long period of time in enterprises. And those workloads fundamentally cannot be repositioned to operate on top of the public cloud. They were designed in a world where the assumptions about what the underlying physical computing infrastructure would do were simply different. There's just a lot more variability in the service levels. There's a lot of things that are different about the way a public cloud provides computing to the user that mean these sort of traditional workloads are going to have to run in an environment where the enterprise has some greater degree of control over the physical computing platforms that they run on. And so that doesn't mean that these enterprises aren't going to want to avail themselves of the cloud to the degree that they can, but it creates this dichotomy that I talked about before where there's sort of a bright line and you have on the one hand the public cloud vendor saying, hey, change all your workloads so they can run ideally over here. And you've got the private enterprises saying, on the other hand, well, I have all these things. I can't change them in the immediate term so I got to keep operating them with them the way they are. And so I've got this bright line and some stuff is over here and some stuff is over there but it creates two worlds. And I think the interesting thing, the thing to keep an eye on about Nutanix is again with some, I think, incredibly complex engineering underneath the covers, they really are looking to offer platform enterprises which says, hey, you can have a single view and a single substrate and you basically can have workloads that run either place and we will take care of the underlying manageability of that and to the extent that the user tries to do things that aren't wise, the fabric will actually be intelligent enough to guide them not to do it. They're making a bet that there's going to be a need for that underlying manageability which is a very good bet. They're also making a bet that predictable workloads will sort of stay on premises and non-predictable workloads will go in the public cloud. That's a little fuzzier. Andy Jassy would have a little debate with that and start talking about reserved instances and all kinds of things that they do around predictability but to be continued. Thoughts on what you're looking at? Classic VC question, what's exciting you these days? Yeah, we continue to be excited about, I think we're in early innings of where computing is going and where enterprise applications around data, around mobility, all these areas are in very early innings and so there's companies like Nutanix, you mentioned a few others, AppDynamics, Bromium, MuleSoft that are, I think, riding these waves and have long-term visions about how they can continue to ride the waves and become very large companies. We see a lot of new innovation, particularly in the area of data and how at large scale it can be used to fundamentally change the way applications behave. Those are some of the areas that we're pretty excited about. You're up to Mr. Ball data, I mean, there's been a little hudduke blowback but I'm inferring from what you're saying that we're just seeing the beginning and it will explode. Yeah, again, I think that's another good example. We believe that there's a set of tools and higher layers of abstractions that are needed on top of platforms like Hudduke to make these things usable to the average user. I mean, Q-Ball's a great example. This is a team that came out of Facebook. They were probably one of the largest kind of Hudduke instances at scale and they said, you know, how do we take what we've done in a place like Facebook and wrap that in a way and deliver it through the cloud so you don't even have to stand up the servers so that your average company that wants to have, you know, use Hudduke in some fashion can just consume it as a service. Yeah, and then insights for the masses, that's the holy grail and we think it will happen. It's just, you know, taking a lot longer than people had hoped but it's complicated as we said earlier. Yeah, I mean, the tagline I heard that really brings it home, the more data I think has been created because of sensors and internet of things and just, you know, the whole big data movement. In the last four years, more data has been created digitally than in the rest of human history. So it's pretty stunning and that's only going to accelerate. All right, Ravi, we have to leave it there. Good time to be a VC, good time to be an entrepreneur. Great to be at the next conference in Miami. We'll be right back right after this word. This is theCUBE.