 Hi, it's March 26, 2014. Welcome to Cube Conversations. My name's Stu Miniman and joining me for this segment is Jeff Kelly and we're coming to you from the Wikibon World Headquarters in Marlboro, Massachusetts. Jeff, some of the news have been calling this Cloud Week 2014, always a lot going on in the big data space, so a lot to cover. Let's dig into it. So first of all, the Cube is today at the AWS Summit in San Francisco, a little bit of news out of Amazon, but the first one, kind of the big announcement this week is Google with their cloud made a big announcement. The Google Compute Engine has a lot of feature functionality. You and I had a chance to kind of dig into some of it there. What was your take on Google's announcement this week? Well, from a big data perspective, they announced a new service around streaming data analytics capabilities along developers to build big data applications to take advantage of real time data. So that was interesting. Kind of look at that as an answer to a correlation to Amazon's Kinesis announcement from earlier this year, or I should say from late last year. They lowered pricing. Google did, so they didn't just lower pricing. I mean, that was a big discussion point, Jeff. I mean, they said slashing prices up to as much as 85% compared to the competition. And not only that, but they really promised that the pricing of cloud will follow the hardware decrease. So really, we should be able to follow Moore's law, which is big if they actually stick to it. Absolutely. There are questions about TCO of cloud deployments, especially when you're talking about big data over a long period of time, whether it really is less expensive to do some of these big data projects, especially when you're talking about persisting large volumes of data in the cloud and doing that over time, whether that makes sense from a financial perspective versus bringing it in house, buying your own hardware and all that kind of good stuff. So it'll be interesting to see how this announcement impacts those calculations. And if it spurs some enterprises, it's going to rethink what the cloud is useful for vis-a-vis big data. Yeah. My commentary on that, first of all, it puts direct pressure on Amazon. First of all, we know if we have short usage instances such as many big data deployments are, Google will charge less than an hour. So that was something that they announced when they came out in public, as well as if you're going to be using it for at least a month or longer, there are discounts that you get, which competes against the reserve instances that Amazon have. I've seen some initial analysis on that that says that maybe if it's a month up to a year that Google is cheaper, but Amazon does have some pricing if you were to lock in for three years, that can make it very competitive there. Yeah. Well, I think it depends on what the workload is and what you're trying to get accomplished. In big data projects, we've seen the cloud, AWS specifically, be a great place to kind of do some of those early experimentations where you've got maybe an idea for big data application. You basically build a proof of concept in Amazon. You don't have to pay for the hardware internally. You swipe a credit card and off you go. The question from there is if it's a successful POC, do you then scale that out in Amazon's cloud or do you bring that back in-house? Back to your point. Certainly, AWS has been a place where we're seeing some of these experimentations happen. If Google can put price pressure on Amazon there, that could have an impact. But I think the area that they really have to focus on now if they really want to get that kind of business is the tools around making it easier for developers to build these applications. As we talked about on theCUBE in past segments about AWS, it's a really, I would say a pretty solid portfolio of big data tools, whether it's Kinesis for streaming data, Redshift, hugely popular, scale out, essentially data warehouse, DynamoDB for no SQL workloads. So they've got really pretty comprehensive offering whether Google can match that. They've got big query. It doesn't seem to me they have quite the breadth of applications. Yeah, I mean, Jeff, I'd agree. There's no doubt that Amazon has a lead when it comes to how many features and how many pieces they have. On the keynote this morning, Andy Jassy spent half an hour building all of the solutions that AWS had. The punditry watching were like, well, there wasn't much new here and we've seen most of it. But if you compare and contrast it to what Google was doing, it's clear that Amazon has a lead. Amazon has a lot more deployments. They have many more customers. They've got tons of revenue in there. And the other place where Amazon really is leading Google is just global availability. They have a lot more data centers. They've built that out. And while Google has a network that's pretty much second to none in the world, they do have work to do to build this out. I'd agree. If you look at Google App Engine has been around for a while now that they do have a lot of pieces to enable big data deployment. I mean, many could say that if it wasn't for Google, would we even have big data? I mean, Hadoop and everything came from MapReduce. Absolutely. I mean, Google is really one of the originators of what we call big data today. They have the original problem indexing the web and storing all that data. And of course, they had to come up with a new way to do it and publish those papers in 2004 around MapReduce and some other things. I think the underlying this whole conversation, though, is the question that others have asked as well is, is Google serious about this? Is Google committed to really offering a enterprise-grade cloud? I'm still not convinced necessarily. We saw, and optics matter in this market, and we saw during the Google event, the engineers and even some of the more senior executives up there in t-shirts, essentially, and jeans. That's kind of the Google thing and the Silicon Valley thing. But if you're trying to appeal to risk-averse CIOs and enterprises in more traditional markets, that's not necessarily what they want to see. So Jeff, to that point, are they trying to reach the enterprise today? Or is this the developer community? I mean, remember, at Reinvent, Amazon gave out hoodies to everyone. Now, they did have their executives on and a button-down cherd and a nice blazer. Andy Jassy would make a great guest on theCUBE, I'm sure. But there is that dynamic. And it leads me to an interesting question. If we look at Platform as a Service, for the first time we heard Amazon today actually talk about the fact that they really are creating a PaaS layer themselves. And if you look at the PaaS vendors out there, most of them want to be independent of the infrastructure layers. So if you take Pivotal, for example, has a partnership with AWS, but you could also work that on Google Cloud in private environments. And I have lots of flexibility. Most of the PaaS providers want to give you that flexibility. And many of them see Google as a good opponent to Amazon to help make sure that there's flexibility and openness. Because if I develop only on Amazon, really that application is going to be stuck there. As opposed to if I use a PaaS, I'm going to have openness and flexibility. Well, so to your first point, do they need to attract developers? Absolutely. For any kind of cloud-based, whether it's infrastructure or more specifically the platform, Platform as a Service, how do you gain traction with developers? And then hopefully that brings you into the enterprise down the road. So very good point, worth noting. I think ultimately, you're absolutely right. First of all, there's some question about what exactly is Platform as a Service, and how do you define that? There's a whole number of questions around that. But if we accept that Platform as a Service essentially is a layer above the infrastructure and below the application, for building applications, essentially. Absolutely, you need it to be cloud agnostic because you don't want to get locked into any one provider. You want to be able to move this between your private cloud, whether it's AWS, Google, VMware, whatever the case might be. So certainly that's important. I think Pivotal should get a lot of credit for what they've done in terms of partnering around Cloud Foundry, bringing in IBM and other potential competitors, at least competitors in their big data side of the business, as they build that out. But still very nascent market. Even there's still, even though, as I said, consensus around definitions, are the database tools that AWS offers on top of their infrastructure, do we consider that Platform as a Service? I would say yes. But there's involves more than that, and it gets into all the tools and developers used to build applications. And the other point on PAS today, of course, is it could fit on many of the clouds, but a lot of the deployments today are going in private environments, in which case some of the points that we raised today are a little bit moot. Last piece I want to talk about Cloud is Google actually, I'm sorry, Cisco made an announcement, a billion dollar investment in what they're calling the InterCloud. Of course, Cisco was one of the original four horsemen of the internet era. Everybody needed to use Cisco. The enterprise uses Cisco, and Cloud in general is a real threat to Cisco. And to be honest, it's a little bit of a weird announcement because some people are saying that Cisco is looking to compete with Amazon, but that's really not where Cisco's strength is. Cisco right now has their partner conference going on. They have lots of relationships with service providers that buy their gear, and Cisco will have a real investment to create services. And there will be some Cisco data centers, but a lot of the service providers will have data centers also. So even with a name like InterCloud, it's really more about the connectivity and portability. They talked a lot, Jeff, of actually about being able to port between various hypervisors, which is a little odd because when I think about most infrastructure as a service providers, while there is virtualization in there, I don't usually have to think about that. Well, isn't that the point of some of these service providers that you don't, they abstract away that complexity? You don't worry about it. You worry about the higher level ability to have applications running your business, essentially, in top of it. I'm a little confused. I defer a little bit to you on this because you're kind of our networking guy, and I don't quite understand Cisco's play here. Well, I think I was reading a report from a financial analyst talking about Google, and if you look at the price we're going on between Google and Amazon, it is the traditional enterprise guys that are under tremendous pressure. So if I'm a NetApp or an EMC or a Cisco, where most of my hardware is going into the data center today, and if that is going to have a significant shift to the large cloud providers, I'm in trouble. If we go back to the discussion we had about IBM, IBM got out of the server business because they didn't see enough margins there, and there wasn't any growth, and they've pushed that all off to soft layer. So cloud is where the margin is, and I think Cisco sees the running on the wall, and they made an IBM move. Here's a billion dollars. We're going to really show you how we're serious here and push forward. Although IBM, I think, committed $2 billion, but that's what's a billion here. And it was very clear when you said, all right, IBM had a billion dollars. Here's the data centers. We're going to build. Here's the services we had, and here's our partner strategy. Cisco is still kind of finding their way in a lot of this. I think what they're trying to say in this announcement is, look, we see the future rather than being dominated by just one or two mega cloud service providers such as Google and Amazon. We see this as, we would rather see a plethora of smaller service providers, maybe specialized in different areas. And if that is in case, in fact the case, as well as kind of private clouds and other internal deployments, well, you're going to need to move data around. You're going to, networking is going to be a critical part of that. So that would make sense to me that they would want to support that vision of the future. It's not clear to me at all if that's going to happen, but I can see why they would want, they would be rooting for that type of outcome. Okay. And that gives us a transition point, Jeff. There's been a lot of funding news lately. So on the last two weeks, Hadoop, big news in the market, big announcements, can you break it down for us? Sure. So last week, Cloudera held their first ever analyst event. In fact, it was the first ever kind of analyst day event from any of the Hadoop providers. So that tells you a little bit something about the maturity of the market. We're getting to that point where they feel like it's important to invite us analysts together to kind of tell us their story in a more cohesive way. So that was good. It was a good event, but one of the things that happened there kind of making it a little bit more of a dramatic day than it otherwise might have been was the announced Cloudera that has announced new funding. We're in our lunchtime that day in front of the team of analysts, the group of analysts there. $160 million in funding led by all people, all companies, T-Rope Price, Google Ventures also included, and Michael Dell's venture arm as well. And then, of course, not to be outdone this week, just yesterday, Hortonworks, competitor to Cloudera, depending on who you ask, raised $100 million of their own. So I wrote, or I should say remarked to some of our SiliconANGLE journalists yesterday that this is clearly, to me, the race is on for an IPO in this market. The two both companies have stated publicly that's their goals to go public. So they're going to use this money presumably to shore up their internal operations, kind of get their ship in order, which you need to do before you can kind of go public. What it does tell me though, especially around specific to Cloudera's announcement, I mean, that's a lot of money, $160 million. They're now about $300 million in investment that they've taken. Pretty much prices them out about $1.8 billion valuation right now. Pretty much prices them out of a potential acquisition. So clearly, IPO is the route they're going. But it tells me that large around at this point in their life cycle that they're not quite as close as we as we had anticipated in terms of when they're going to go public. It could be more than a year out. I think a lot of us are thinking it might even be this year or early next year. I think it's going to be a little bit further out than that, because this tells me they need some of that funding to, as I said, get that, get their finances in order and some of the other internal things you've got to. And then the other thing that I learned at that show or at that event, I should say, is from a customer perspective, they've got about 300 paying customers, they said. I would expect that number to be bigger at this point. So for me, this is really all about using this funding to scale their operations, scale their customer base, drive some real revenue, cross $100 million revenue annually, annual mark. How far are they off from that $100 million mark, which is usually kind of one of the barriers? Yeah, I think, so we estimate this year they were between $70 million and $75 million in revenue for 2013. So I think it's doable that they'll get to that $100 million number this by the end of this year. But the question is, and what got me thinking about this, to some degree, was the Box IPO and the S1 talking about how much money they're burning through spending, lost, I think, $168 million, and that's with $120 million in revenue. So I don't know the necessarily the exact amount of cash clutter is burning through, but I would suspect not only do they have to hit the $100 million mark, but get within a stone's throw of at least cash throw neutrality. Whether they're there yet, I don't know. I think it's going to take a while. I don't think they're there yet, I should say. It's going to take a while to get there. So while we're talking funding, another $100 million round raised by a local company, Actifio. Actifio, Jeff, if you haven't run across them, they're really looking to, they're in the backup and storage world and changing the way that's done. So similar companies, people that are pretty much familiar with data domain, Actifio has a new way of doing kind of a copy management, because one of the problems we've had in storage for a long time is you make copies for lots of different reasons and they end up all over the place. And Actifio is a real simple way of making that more streamlined. They therefore use less storage and it's been pretty popular. They built a next new headquarters in Waltham, Massachusetts, down the road from us here, and $100 million. So good proof point that there is still a lot of excitement going on in the storage world. They are primarily a software-based solution fitting into the Hall Software Defined Data Center. Discussion there and interesting one to keep an eye on. I'm sure they're driving that towards an IPO. Well, yeah, obviously there's a lot of action happening in the venture capital world. It's an exciting time to be in this market, to be analyzing this market. There's companies in both storage and big data and analytics that are really trying to kind of revolutionize the way companies manage data, make use of it, store it. And it's really interesting and clearly there's a lot of money being invested in big bets being made. Yeah. One last one I'll mention is in the networking space, a smaller company, Embrain, got a $14 million round here. Embrain is one of the companies that I heard a bunch of people talking about at the Open Networking Summit that we covered a few weeks back. Application-centric networking. So really, we talk about software and how that's leading the change. In many ways, it's changing the management layer of networking as well as enabling various services to be able to be offered in a software fashion. Embrain is a partner of Cisco's. While there's many companies out there that see SDN as really the disruptor to unseat Cisco, Cisco does have such strong power in the marketplace that it's still important that you're going to fit into that ecosystem. Yeah, that's an interesting point, not specifically around networking, but just in general about all these new approaches to storage, networking, big data. Are they disruptors in the sense that they're going to really disrupt the business model and the technology of existing incumbents, or are they going to be complementary? There was a question around that in Clodara's event. They stressed numerous times that they're not trying to replace, for example, the traditional enterprise data warehouse. And I agree with that, that they're not trying to replace that. They can't kind of do some of those workloads that data warehouses do. But clearly, you're seeing there's an overlap in terms of functionality, and clearly, they're competing for the same dollars in some cases. So if not replacing, they certainly are competing. And I wanted to clarify that point because I've been quoted earlier in some of the media reports out there about the fact that Clodara was trying to compete. And I don't think that's quite what's happening. It's much more subtle than that. But I wonder if something similar is happening in the networking and storage space. Are some of these disruptors trying to replace EMC and NetApp in the storage space, or Cisco in the network space, or are they trying to comment? Absolutely, Jeff. So if you talk about potential IPOs, it's widely expected that Arista Networks will IPO this year. They are a classic, direct competitor to Cisco. They looked at, if you look, Andy Bechtelstein, Help Fun Son, and J. Shreya Lal, who was a longtime Cisco executive, found this company. They had a new way of creating networking. They have really an open format for being able to program and create their switches. So for high speed, especially low latency environments, they fit well. They started in a niche environment, especially in HPC environments, and have spread into the enterprise, and have slowly been chipping away at Cisco head-on, as opposed to some of these software-type products, which might fit in with either Cisco's ecosystem, which they're building with the application-centric infrastructure, which is ACI. Others are rallying around what VMware is doing since the NYSERA acquisition, and trying to have much more of a software-driven, let's commoditize the hardware piece, and therefore pull dollars away, because, as you alluded to, Jeff, for many of these spaces, there's only a limited amount of dollars that companies can spend. So if we can take 50% out of what the incumbent is charging, there's more opportunity for a number of them to play in that. And of course, a big area that I cover is the convergent infrastructure market. So if we can take the whole bucket of compute, networking, and storage, lump it all together, do it at a lower price point, improve the operational efficiencies, there's a lot of opportunity there. Absolutely. Well, again, interesting markets, I should say. So we'll be watching closely. Yeah, so the last thing to wrap up, Jeff, we've always got a lot of events going on. If you check out siliconangle.tv, you can actually watch right now where we've got a big cloud tour 2014 going on. Amazon summit in San Francisco is going on right now. Next show that I will be out personally will be the Red Hat Summit coming up in mid-April. And, Jeff, how about on the big data tour? Yeah, we're excited about our next event, Sports Data SV. So we're going to be broadcasting live from AT&T Park, which is the home of the San Francisco Giants. This would be a great backdrop. Actually, I believe this will be our second, maybe even third event there at AT&T Park. But this one's going to be focused on using data to essentially the use of data in professional sports. We're going to have guests from the Giants, from the San Jose Sharks, and some of the other local professional sports teams, talking about how they're using data to improve their business, to improve the team on the field or on the ice, as it were. Yeah, moneyball type stuff. Moneyball type stuff. And we're going to question some of those moneyball assumptions. And how much of moneyball is real and how much of it was literary license, if you will. So it'll be interesting to see how real world sports franchises are really using data to run their business and become more competitive. All right. Well, Jeff, it's always good to catch up with you on everything that's going on in the space. Thanks, everybody, for watching this episode of Cube Conversations. We'll be back for more plenty soon. And don't forget to watch SiliconANGLE.tv for the live event, wikibon.org, for the research. And you can hit us up on Twitter. It's Jeffrey F. Kelly and I'm Matt Stew.