 Live from Midtown Manhattan, the Cube's live coverage of Big Data NYC, a Silicon Angle Wikibon production made possible by Hortonworks. We do Hadoop and WAN Disco, Hadoop made invincible. And now your co-hosts, John Furrier and Dave Vellante. Okay, we're back. This is Dave Vellante and this is the Cube. We're here at Big Data NYC at the Warwick Hotel. We're winding down day two. The Cube party's just about to start at six o'clock. So if you're around the neighborhood, come on in. We're at the murals room. I'm here again with two folks from WAN Disco. David Richards is the CEO and Jagain Sandara's back. And we're going to talk a little bit about the industry, the road maps and futures. Gonna take a business view and also a technology view. David, I want to start with you. We were talking off camera. We have all of us seen all these shifts and bubbles and momentum moves in the marketplace. And it feels like there's another one coming, a lot of disruption going on, a lot of innovation. We've seen newcomers, we've seen existing players, making moves, buying companies. What's your take on what's going on here? Is this a bubble, is it a new marketplace? What's your take? So naturally when we see disruptive technology, disrupt a marketplace and begin to take huge trends in market share, we have a natural tendency to call it a bubble. I don't think it's a bubble. I think this is a trend, a very clear trend, and we're deeply ingrained in the fabric and infrastructure of applications. We talked earlier about the average life of an application in the enterprise today being 19 years old. We're now asking people to rewrite those applications on this new platform. It's called Big Data and the underlying infrastructure behind that is Apache Hadoop. We were just talking about what big, how big companies react to new trends and the latest fad in technology, and we talked about the rice principle, right? And I won't name names, I won't mention or at all. But me. Sorry. But I think we do see the same thing. History has a habit of repeating itself. Rice principle, first of all, they ridicule, then they ignore, then they criticize, and then they embrace. And we're just beginning to see the first signs of some of these larger infrastructure technology companies realize that Hadoop is probably not a fad, it's not a bubble, and they're beginning to embrace that technology by building applications that integrate with it. I think this notion that use Hadoop for cheap storage and then if you wanna do something really useful then move it into some relational database somewhere, I don't think that's quite right. Saw it from the East Coast and I remember well the mini-computer debacle, the digital prime, Wang, DG, on and on and on and on, sort of ignoring the PC trend. It seems like the marketplace today is quite a bit different. The leaders, many of the leaders of companies, whether it's Intel, Microsoft, Oracle, EMC, others, created a lot of that disruption. So they're well aware and they're paranoid to use an Andy Grove phrase. And so that's sort of one point. The second is that the market's shifting again. So it used to be IBM vertically integrated and then competition was created on long layers and you had leaders emerge in each one of those discrete segments. It feels like it's coming back together in a sense today either with vertical integration or with ecosystem integration. What do you take away from the structure of the industry and the right structure to compete? So I think the shift in the marketplace, as you were talking about sort of microcomputers and missing the whole PC thing, I think Hadoop is open source, right? Open source has a habit of being highly disruptive in the market to a large degree and it shifts and makes changes at 100 miles an hour. It's a very efficient distribution mechanism and kind of overnight before we even realize it, open source finds its way into the fabric of organizational infrastructure. Techies train themselves on how to use it. It goes very, very quickly to the heart of enterprise. And I think unlike proprietary technology which kind of weaves its way in very, very gradually, open source kind of happens like the Linux movement happened overnight, happened very, very quickly and the shifts tend to be much more dramatic, much more sudden and consequently very, very unexpected. And I think that's what we've seen. I mean, it was only three years ago, right? The show was tiny, you know, you probably had a couple of hundred people here. Now there are thousands of people here. It's happened overnight and that's what open source does. It shifts and changes markets very, very quickly. Yeah, it's like a small lever can move a big boulder type of thing and it is very unpredictable. Like we were talking earlier, but I was using the example of canonical and the disruption that it's having on in the Linux marketplace in general and specifically with Red Hat, even though there's really no money being made there. And yet it's a major force in the marketplace. So again, I want to talk about the technology, you know, trends that we're seeing, you know. The microprocessor revolution seems so trite now. The shift, but at the time it was massive. I mean, people forget IBM at one point had 50% of the revenues of the computer industry and two thirds of the profits. That's a monopoly towards what Microsoft and Intel had. So what are the major technology shifts that you see enabling these changes? So the first thing I'd like to say is that this technology is real. The scalability and the cost considerations that Hadoop brings are well proven. The pieces that I'm missing are mostly enterprise enablement focused, so it's not a bubble in any sense of the word. The second thing is that much like we moved to personal computers, this move to clusters of computers running either in the cloud or on premise is a tectonic shift. It's going to move from Oracle Rack, say supporting up to 64 nodes, which was the highest scalability they could reach, to a multi-petabyte cluster of Hadoop nodes, a few thousand nodes. That's a tectonic shift. Do we need this technology? Yes, the data that we're processing is so much more vast and the fidelity you get from the queries you make is so much better with this volume of data. So the technology is there, the need for the technology is here. I believe that this is truly a tectonic shift. The existing players will do reasonably well. At the end of the day, they have businesses built in servicing enterprise customers, so they will do reasonably well, but make no mistake, as David pointed out, the open source wave is upon us and it will change the way that enterprise vendors do business. And the economics, right, are radically different. It is radically different. If you consider the cost of storage in Hadoop versus the cost of storage in traditional sand or nas environments, it's hardly a fraction. It's five to 10% of the cost of nas storage, so it's dramatically different. As a technologist, when you look at the mobile trends, 15 billion smartphones at some point in time, what do you think about and how do you think about that impacting your business? So the first and most visible impact is the volume of data produced and that just is multiplied several orders of magnitude, so you need big data systems. The second less visible aspect is that the low cost computing that the microprocessors built for cell phones bring is actually beneficial to huge servers as well. There is a notion of running arm type servers in Hadoop clusters that's gathering momentum. Cost savings from saving energy is real, so I think there are those two aspects of mobile computing's impact on big data. So David, I want to come back to you in terms of just, again, trends in the marketplace, companies that are formed. So we saw VMware, small company, just totally disrupted. You're seeing companies like Workday seem to be on a rocket ship, ServiceNow and Cloud, Splunk, Tableau. There seems to be this renaissance in software. Now, VMware, of course, got acquired for 635 million, probably the greatest acquisition in the history of the computer industry. Now it's got probably a $35, $40 billion market cap. These companies tend not to be able to stay independent although I guess Workday probably will be able to after the poison pill measures that it's taken. But it's still tiny, relatively tiny today. Do you feel as though that a new, let's say billion dollar, I mean I guess the last one really the Mirain Independent was Salesforce, right? But is there an opportunity for a new billion dollar software company to emerge out of this big data trend? I think just like everything in life, we work in cycles. And I think back to the original comment that I was talking about applications being sort of almost two decades old, I think this is a new cycle. And just like a decade or so ago, two decades ago almost SAP was minted, JD-Ed was minted, PeopleSoft were minted. These large companies were created out of the trend called ERP, Enterprise Resource Planning inside the enterprise. We are going to see a new set of companies up and coming, Cloudera, Hortonworks, Pivotal, who I think could all become enormous companies in the size of those that you mentioned earlier. I just think it's a natural cycle that we're going through. I've got no doubt that all those companies I just mentioned could have been taken out if they wanted to sell. I don't think that they want to. I think they see sort of a $50 billion marketplace and they can remain independent. They can probably eventually go to the public markets and raise capital there as well. There's no reason for them to say, why would they sell? Well, I mean, there's no doubt that Diane Greene must look back at selling 635 million. She sold out way too low and got thrown out of the company. So I mean, a lot of people have seen that example and said, okay, we're not doing that. On the flip side of that, many say, hey, let's just get out, you know, get in, get out, we're in it for the flip. And you've got big companies that have huge cash hordes. And what are your thoughts on R&D? For these big companies, it seems as though acquisitions are a lower risk use of capital than R&D sometimes. Well, yeah, I mean, there's a notion that really Silicon Valley is one huge R&D center funded by Venture. And I think there is something to be said for that. And niche plays undoubtedly have to be acquired, right? If you're in a small niche of a much larger marketplace, then acquisition is a natural exit. And what you guys did, it was genius. I mean, it just exploded you into the business. I mean, it was magic, right? Imagine if you had to develop that on your own. Well, it was, I mean, for those of you that don't know, WANDISCO started and maintained itself without any venture capital. The first time it raised any money was going to the public markets to do this thing called an IPO, which most people thought this is highly improbable and just impossible for that to happen. But the great companies are actually born, I believe, out of building products for necessity, necessity being the greatest mother of invention, right? So we were always building products that we knew enterprise needed to adopt. And I think it created a great discipline inside the business. Now we're unusual, not every company can do that. Certainly when a lot of the big data guys that are trying to create a market here is very, very, very difficult to do it without venture capital. So, you know, queued us to those guys that have had to go and raise venture. We very fortunately managed to do it without venture capital. But you think that these emerging companies, many of the ones you just mentioned, Cloudera, Hortonworks, et cetera, can remain independent. Yeah, I do. Whether or not they want to, that's one thing. And you can question it, but I agree with you. Many of them do want to remain independent. And you feel like they can become acquisition proof. I think they will probably look toward the splunks and the tableaus of this world and the work days and look at the multiples that they're getting in the public markets. And if I was a shareholder in those companies, I'd be really saying, you know what? If we exit now, this could be another VMware where we're going to exit for a 10th of the real value. I think they're going to have to push as far as they can possibly go and really go into the public markets and see what life's like there. Well, on a short-term basis, the shorts were out earlier this month on Tableau. I mean, they were just really vocal, pounding the stock down, and I shook my heads and this company's crushing it. They're going to have a huge quarter. The shorts are going to be running for the hills. It's exactly what happened. But longer term, I feel like these companies aren't even scratching the surface. I feel as though this whole notion of analytics for everyone is something that is a very powerful concept. 2014 is going to be the year when big data moves hugely into the enterprise. We're seeing so much movement where so many companies are either, everything from manufacturing companies, Telco, banks, of course, who usually move first in these new markets. All of them are staffing up big data right now. Both in data centers and people. That's a surefire sign that we are going to see huge moves in this market. And you get a lot of infrastructure works being done to harden the top, as John Furrier sometimes says. You guys are playing a critical role there. But that notion of analytics for everyone, there's new technologies that have to be developed. Software, user interfaces, the ability to automate the cleansing and preparation of data. I mean, it seems like there's an unlimited amount of white space. What are some of the things you're tracking there? So our own company's core intellectual property, of course, is continuous availability. And that's the first step. You've got to have the data available at all times, at all places. And that's what we do. There will be security place in this space where access to data is much better controlled, final-grained security. There will also be in-memory analytics place because there's a lot of data to be collected to collate this vast amount of data into something that's consumable by human beings needs to have an in-memory step in the middle. Aside from that, if you really look at it, the businesses that don't adopt big data and use it to drive decision-making will be at a tremendous disadvantage. So you'll see those fade. In the natural evolution of businesses, those will fade and will be supplanted by very big data-savvy competitors. And that's where I think we, as technology companies, need to provide technologies that enable that. Where were you guys seeing traction? I mean, obviously, we're in New York, financial services, how about service providers? I mean, are they sort of knocking your door down? Talk about that a little bit. We're seeing a lot of demand, particularly from financial services, to be quite frank with you. They understand that if they're gonna deploy Hadoop in their enterprise, it has to be continuously available because they're usually using it for mission-critical data. So it's not even a might-need-it-some-must-have. So clearly that's a place. Telecommunication industry as well. We've seen some very, very, very interesting use cases in telcos where they're moving, we announced the deal with NSN, which is an OEM deal, which is a replacement of traditional database. And there, we're seeing significant demand and continuous availability there as well. And service providers too, I think we've seen a couple of interesting instances. Some of the cloud guys are looking at this who are deploying Hadoop for the cloud applications and have to have continuous availability where the data center isn't a single point of failure too. Yeah, I think that as is often the case, financial services sort of leads the way, it leads the charge in situations where you need mission-critical high availability, continuous availability, or some variant thereof. And then the rest of the market learns from that and begins to expect it. And it seems to be happening faster in adjacent areas this time around, because of cloud, because the service providers are getting massively disrupted and trying to race to compete with the Amazon effect and the Google effect. Exciting times. It is, it's just phenomenal. And just one other trend, it's probably worth mentioning, that if the concept of memory becomes not a limiting factor, why wouldn't you run all of your applications in memory? So I think this is one thing that we have to keep a very close eye on over the coming years. As hardware becomes less of a limiting factor in terms of memory, why wouldn't I run everything in memory? I mean, petabytes and petabytes of data potentially in memory. Well, you know too, when we're talking about the 20 year old app, 19 year old on average applications, every one of those applications are 99% were written assuming there's a spinning media that's going to persist the data. Now you take flash as a memory extension, it just completely changes the way in which you think about developing applications so that that constraint is removed. Oh yeah. Now the network becomes the constraint which is music to your ears. So. Ha ha ha ha ha. You said it no way. All right gentlemen, well thanks very much for helping me wrap up today. Really appreciate it and thanks for all the support. You guys are great analysts. If you ever want to, you know, get out of the technology business and get into the analyst side of it. Hope to do more with you. So thanks for coming on. Let's get up a glass of wine. All right, let's do that. All right, we're off to the CUBE party. We'll be back tomorrow morning at 9.30. So please check out siliconangle.com, siliconangle.tv, and wikibon.org for all the research. Thanks for watching everybody. We'll be right back. All right.