 Hadoop Summit, this is SiliconANGLE and Wikibon's exclusive coverage of Hadoop Summit. This is theCUBE, our flagship program. We go out to the events and extract a signal from the noise. I'm here to join my co-host Jeff Kelly. Jeff, welcome to theCUBE. Scott, welcome to theCUBE. Great to have you here. So you kicked off, helped kick off the show this morning with your keynote. Talking about a number of things, among them the new Teradata clients for Hadoop. Brought it on stage, which I thought was great. I love some. I was joined by a dancing appliance. Wasn't that great? It was fantastic, and a good-looking appliance it was. But why don't you tell us a little bit about yourself, kind of your role, and then we'll kind of get into what Teradata is doing here at the show and some of the strategies you're taking towards the big data market. Okay, great. Well, I'm Scott now, I'm from Teradata Labs, and Teradata Labs is actually an organization within Teradata that is responsible for research, development, engineering, product management, product marketing, all the products, all of the knowledge that we roll out, kind of the innovation engine of Teradata is what we're responsible for. And we've been obviously affiliated with Hadoop Summit we were here last year. It's really great to be back. Having been in the data warehouse, big data, kind of data analytics business for a long time, the one thing I have to say about this whole movement in the Hadoop space is that it's unlike anything else I've seen, in that it's every geography, it's every industry, and there's so much energy and emotion around it. It's unlike any other transition that I've seen. And even the difference between our visit here last year and this year, where we've seen the promise turned into reality, where we've got customers who are implementing, where we've got businesses who are driving value from the solutions that they're really, that they're integrating with the solutions that they've already got. And being able to demonstrate that value really emphasizes the importance and I think will help to continue the momentum that we feel in this market. Scott, one of the things I want to ask you was, obviously the theme at Hadoop was offloading data warehouses with Hadoop's a benefit there, but you have a relationship with Hortonworks. And we were talking earlier with Murph, who was an analyst with Gartner, we were talking about the early adopters and the mainstream getting it now. But there's always a question of value, right? Where's the value? Because there's legacy involved, right? So most of the web-based companies are going to be cloud, they're going to be SaaS, they might have a green field, clean sheet of paper to work with on big data. But in the existing enterprise, large financial institution, insurance company, or what have you, they have legacy technology. And they have to, but they want Hadoop, they want to bring it in. When you talk to folks out there, what are some of the challenges and opportunities they have with that environment and the technology specifically? Sure, that was like a long question. There's a lot of threads in there. I want to really try to hit on a couple of important themes because you hear it here, I get asked a lot about it. One of the things that people often say is, why are you here? This whole Hadoop thing is offloading data warehouses. Isn't that bad? Doesn't that bother you? And the answer is absolutely not. Certainly there's some hype around that and there's some marketing around that, but when you really look at the technology and the value of what it brings to the table, it's a new technology that really allows us to harness new kinds of data and store those new kinds of data in the native format. And storing detailed data in the native format really enables the best world-class analytics. We've seen this happen for, as long as my career is in the traditional data space. So that's a really good thing. The way I view it though is sure, will some workload move around the infrastructure from the data warehouse to a Hadoop cluster? Potentially, right? And by the way, if Hadoop is a great solution for it, it should go there, right? But at the same time, there is more demand than there is supply of technology and what I mean by that is the demand for analytics is so extreme that actually adding this tool to the toolkit gives customers more choice and gives them the opportunity to really catch up with the backlog of things that they've wanted to invest in over time. And then the final point really, I view what's happening here as perhaps one of the single largest opportunities for expansion of the role and size and scope of the data warehouse in an enterprise, because one of the big things that Hadoop brings to the table is a whole lot of raw material, a whole lot more data, data that used to be thrown away, data that never existed a year ago, is now going to be able to capture, be captured, be stored, be refined, be analyzed and as companies start to find relationships, as companies start to find actionable tidbits from the analytics in this huge source of raw material, I think it's actually an opportunity for upside, for them to integrate more data into their data warehouse where they can actually do the real-time interaction and streaming that's going to get them to the demonstrable business benefit. So it's the modernization of the enterprise? It's modernization. The way I look at it is also it's, sometimes the word incremental can be, it can sound like we're trying to downplay it, but I see it as incremental in that it's different data and it's incremental data, it's incremental subject areas, it's new stuff that's going to come into the environment and based on what we've seen in the history of analytics, there's no end to the value that companies find and there's no end to competition in their businesses. So this is a huge opportunity for the entire community to deliver more analytics and I think that there's actually more upside for traditional legacy data warehouse vendors than there is anything. I think that's a really important point because as you said, a lot of people think about that offloading workloads, but it's also about offloading workloads but bringing in new data, doing more analytics and then moving some of that into, back into the data warehouse so you can actually create more value from it. Yeah, I mean, one of the things that I've seen is over time and Moore's law is something that's been going on for some time, right? And cost erosion in hardware has been going on for a long time and you think about the thing that you buy today for your BI implementation, the hardware cost, what? 20% of what it cost three, four years ago. And you know what? Revenues continue to increase because there's such pent up demand that as it gets less expensive, it becomes more consumable. And I think the same thing is really going to continue to happen as we add in these new technologies and these new data types. So one of the things I want to commend Teradata for doing is focusing on kind of that reference architecture and helping customers understand how this new technology of Hadoop and big data fits in with everything else that they're doing. Talk a little bit about how, from a reference architecture and then maybe even from a product perspective, how Teradata goes about turning this into a reality for enterprise customers who really, they're not looking to just kick the tires of Hadoop, they want to use this to really support applications and workflows that are really critical to their business. Yeah, I think one of the biggest things that we can do to help the industry and to help our customers really is to define a realistic roadmap that's consumable for them in their enterprise. And so while it's certainly easy to have a marketing release or a press release that says, oh, this new technology does everything, it slices bread, it washes your car, it does all these things. In reality, there are very few things like that in the world, right? But the new technologies and the new innovations really do fit into some very interesting new use cases. And so by providing this integrated roadmap of how customers can deploy and fit these technologies together is a really great education process and it's been extremely well received by our customers and prospect. I have to tell you that even in advance of the announcement of the things that we had here today, we've already got customers who have gone down this path with us because it's such a compelling value proposition. The other thing is that we don't actually put specific technology in those boxes. It's a reference architecture. We hope that there's some teradata product in there, but at the same time, our customers understand that there is choice in the marketplace and the best solution is going to win. And by providing this reference architecture, I think we help elevate ourselves to more of a trusted advisor status with the industry and in how we see these things fitting together and providing very effective, very low risk kinds of solutions. Well, I think you hit on something, the trusted advisor. I think companies and enterprises are just crying out for some leadership and to help them really understand how they're going to make this a reality in their organizations. And you mentioned kind of the openness and being able to allowing enterprises to choose the technology that fits the work case. Of course, you know, you hope that's teradata in a lot of cases, but it could be something else. So talk a little bit about your relationship with Hortonworks. So I know you announced today kind of a reseller agreement and you're going to be actually reselling the subscription service to Hortonworks service offering. Talk about that a little bit. And also, I want to dive into the tech as well. The Hadoop appliance we mentioned earlier that you announced and maybe just kind of walk us through some of the news today. Sure. So I mean, obviously we have a strategic relationship with Hortonworks and it's our second year here at Summit. And it really started with, I think, a very common view of what's happening in the marketplace and how these technologies should really play well together. At the same time, we also really believe that it's important that the community embrace the open source Apache version of the software so that it doesn't become fragmented and become obsolete, right? So Horton is spot on in terms of business model and putting everything back into the Apache open source version. So that means that I think this is the version that will win and this will be the version that companies can count on to be sustainable. So I think that there's an advantage there implied. So that said, I think it fits into the right place. We've got a great engineering relationship and a great common vision on how the enterprise architecture and how the pieces can fit together and be optimized for different workloads, for different service levels and for different applications. So having that common vision and kind of, I think, bringing two best of breed providers together with Hortonworks on the Hadoop site and Teradata for what we're very well known for, I think it's really the best of all worlds. And we work together to lay out this reference architecture. And so it's not just, you know, Teradata came down from the mountain said, this should be your reference architecture. We got some validation. We got some validation of use cases. And then we went to work from an engineering perspective on how we go build these things out and make them work and optimize them and support them end to end. Because obviously not only with all of the new solutions, is there kind of a scarcity of talent and some confusion, support becomes really, really important. So one of the things we added to our portfolio that we announced today is an expanded relationship on the support side, where customers can come to Teradata for integrated support of all of their data analytics environments, whether it be Teradata, whether it be Aster, whether it be Hadoop with HDP. And you know, that's a really nice thing where there's one phone number to call. We've got fully integrated processes. We can help with a global footprint in the 80 countries where we do business. And obviously Horton works with the extreme depth and ability to manage the content of the kernel can get it done unlike anyone else. Scott, we've been talking enterprise grade all morning. I'll see what that was the theme of the keynote. We're from our garden talk about security, compliance. I mean these are meat and potatoes enterprise issues, right? So I got to ask you, what are you guys looking at? What's coming next? Obviously the platform has to stabilize, developers going to want to program on it in different environments. But the reality in the enterprise is there's certain requirements. So what are you looking at in the labs that's coming around the corner that's going to be really, really important for customers to realize the value of scaling and harnessing the big data of Hadoop with the existing infrastructure? Yeah, I mean I think there are two things that we'll continue to do. One is we'll look to build out kind of that framework of ecosystem. And in all of the keynotes this morning, everyone talked about the value of the ecosystem and it's amazing that ecosystem, how they're just more and more logos this year than there were last year. And I think that that will continue but really building out that ecosystem so that those things that are important can be realized and they can be realized in a very repeatable fashion. I think in addition to that kind of ease of use, right? Because despite the fact that we have burgeoning numbers of newly minted data scientists and people getting into the marketplace, that's really good, there still aren't enough. And so de-risking things by making them easier to deploy and easier to support I think is a key focus area. And then finally, I said two things but now third, finally we'll continue, I know, sorry, we'll continue to look at performance and just making sure that we have the best density, the best performance, the cost performance value proposition that our customers will want because I also continue to believe that the supply of data will outstrip any customer's ability to invest in infrastructure. I'd love to get your take on, I want to go back to what you mentioned about the, you know, the Hadoop distribution focusing on Apache and really being Apache compatible. So I take that number one to me and Teradata is not going to be coming out with their own Hadoop distribution. Absolutely not. But how do you see this? Yeah, I think we can say that pretty definitively. So, but what about, how do you see this whole Hadoop market playing out? I mean, you've got Hortonworks, Cladera, MapR and some others. How do you see this playing out in the next year or so? I mean, is this, you mentioned, you think, again, that kind of the open source Apache version is going to kind of win. When do you think that's going to happen? You've got some competitors in the market and different business models. Just I'd be serious here. You know, there are different business models and different innovators and, you know, my crystal ball is probably only about as clear as anyone else's. But, you know, kind of for the long term, I think it's best for the industry if it mimics a model similar to the way Linux is deployed where there's kind of a duopoly, maybe three vendors. It's very largely open source. There's a lot of portability between. I think that really strengthens the position of Hadoop as a core technology and foundation for some of the things that we're doing. And so I would hope that the most successful outcome would be that we'd end up with a duopoly or maybe three kind of providers around a similar kernel because that would remove fragmentation from the market. By the way, I think we are a software company, so I think it's fair for companies to have value add proprietary software. That's not a bad thing. But at the file system level, at a core tool level, I think the open source community cannot be out-innovated, right? And so I think that that's a really important thing. So I think, hopefully we'll get to that duopoly or maybe three companies that kind of have that. I don't know if we will, but I sure hope we do. And I think if I were to bet on it, I would say it's odds on that that will be the case. Now, will that be 18 months, three years, five years? I don't know. Scott, thanks for coming inside the QBOS. So you guys have a great position in the marketplace and the enterprise message is strong here. That's what the demand is. We're seeing a lot of trends out there that want the enterprise grade, big data, which is not just once there's a big part of it. Thanks for coming inside the QBOS and sharing your perspective and what you got working on. Certainly having the new products come out to be great. So thanks for coming onto the QBOS. This is SiliconANGLE and Wikibon's coverage of Hadoop Summit. We'll be right back with our next guest after this short break.