 Okay, we're back here live at Stratoconference. This is Silicon Angles exclusive coverage of Riley Media Stratoconference in the heart of big data innovation in Silicon Valley and Santa Clara, California. This is theCUBE, our flagship program and throughout the event extract the signal from the noise. We've been at all past three years of SAP Sapphire and we're going to talk with SAP today. I'm John Furrier. I'm Dave Vellante at Wikibon.org and this is our in memory segment. We're talking about the application of in memory architectures, particularly in big data generally and Hadoop specifically. We're here with Chris Hallenback who's a data warehouse architect for SAP. Chris, welcome to theCUBE, but thanks for coming by. Thank you very much, good to be here. Yeah, so we've been talking a lot about in memory. You guys had a keynote this morning. We were at Sapphire as John said the last three years, two years ago was really the big Hannah push as you might recall. I think Bill McDermott maybe even said big data once two years ago in his presentation. Last year it was only up to maybe 10 or 15 times. You guys haven't been over the top with big data washing. I'll give you credit for that. At the same time, I think Jim Hagam and Shinabe came on theCUBE and he was saying, yeah, you know, big data. We're really about fast data. John Furrier coined big fast data as SAP, but so now however, you guys have really caught the wave, you're hanging out to Superman's Cape and jumping right into the big data. So give us an update on Hannah. What's happening here? What are you guys doing at the show? Well, I think Hannah's doing phenomenally well. We just came off an amazing year with just fantastic revenues or ecstatic, but more so it's been the adoption side and the various use cases. Because you can't really take the technology and then just say that it's completely new that does new things and say, here's going to solve this problem. You need to come up with problems and then people find the right technology for it. And those situations are now happening. How can you take a system that can take, you know, billions of records an hour and do advanced analytics on it? How can you do, and I don't actually even like in-memory because I think it's focusing on the technology, in-memory is the capability along with a lot of other deep, deep work that we've done. Well, you have a hassle to blame for that by the way. And we do. But to get to- Because that's all he talks about. But to get to real time, how can you actually take a system without doing huge amounts of data preparation? How can I take data in, stream it in real time and do analytics on it at that second, right then, as well as dealing with different formats? I think people focus in on relational and clearly we can handle table data. But then what's unique about HANA versus building these complicated landscapes that we've gotten through NoSQL and other stuff is that HANA also can take that data that's accessed relationally and make it available multi-dimensionally with just the metadata layer to do that. We don't make a copy of it. So now I'm getting the same access speeds as a multi-dimensional database that took hours and hours to roll up. I can also then deal with big titles. I can do dynamic columns on the fly with big table implementation. So I can have a fixed table and then add columns dynamically on the fly for my marketing problems without having to work with engineering to pre-do that. We're getting into adding spatial analytics to the system. We also have extremely deep text analytics. So I can take unstructured information. I can do language identification, stemming and tokenization libraries on the main 10 languages. So I can do deep sentiment analysis all in HANA in real time on data without data product. Yeah, so I think that you're making a really good point. We all, we like to talk about the tech but it's all about the business value and the productivity impact that systems like HANA can have on an organization. I'm struck, I remember, I think it was last year at Sapphire. I think it was the CEO. I don't think it was the CIO. I think it was the CEO of McKesson who was up talking about HANA. So you get, you know, SAP, you guys, you know you're all about business. So we're touching a lot of different parts of the organization with this. Well, and that's what we do. We go, because now, sure, I'm a web startup and I can deal with a complex landscape with lots of different layers of caching and memcache and I can run graph database over here. But when you're at a big, when you're at a corporation, you want to run this stuff and you're at McKesson. All of a sudden, I can't have differences admins for everyone, every different system out there, every database, everything. I want a system that does it, that has all the different functionality I need but allows my engineers to write in the right language whether it's a multi-dimensional dialect, SQL for single updates or something that allows me to get at other information that's stored as a document and I want to use JSON. And so HANA enables these developers to be incredibly efficient inside the organization by writing in the right paradigm that they want but only having one database to manage. So I don't need to have all these different specialists on all these different systems out there and those SQL variants, which are, many of which are awesome but many companies don't want all those different pieces. They want one system that solves those problems and they can do that all in real time. But now our previous guest was talking a little bit about the greatness of SAP as an ecosystem player, you guys as an application provider. He's a small SQL database specialist and he expressed frustration that HANA is basically locking him out. He's closed, he can't bring his database and his use cases and his small customer base and take advantage of HANA. What's your response to that? SAP's philosophy, will you open that up for other, sort of no SQL databases? Why or why not? What's the reasoning and the philosophy behind the current posture? First of all, establish what your policy is there. It's funny I'd say it's totally opposite because coming from, I mean, ours actually to embrace other databases. So a lot of the work we're doing. I mean historically that's the case, right? Yes, we've opened up our APIs to third parties that are directly competing on our BI layer. We've opened it up, I'm working directly with a lot of independent software vendors, ISVs to port their applications to HANA that make exact copies like competitive products with our own and now we're opening it up, eventually moving into, we've talked a lot about starting to work towards data federation and other pieces that will tie into all the other databases so that you can work with HANA and you can get data out of other systems or allow them to get data out of HANA. So we're trying to build and be extremely open in our ecosystem. Now I think the part of that might have been more on the application side and clearly there is when you're writing an ERP system you need to be on an ACID compliant system that can do transactions and handle that type of load and that's a different channel obviously. I'm on the database world and I think in that world yeah the major players are out there that you have the oracles and IBM and obviously our own Cybase ASC and we've extended that to HANA but that's a lot of you verify. But just to clarify, so the effort these approach is open through your APIs, I can bring any NoSQL database into HANA and apply whatever use cases I want from my customers. And it's a matter of just interacting with your API. Yeah, and our APIs are open. All right, we'll see you here or here. So now I want to talk about Intel a little bit. You guys, you saw the Intel distro announcement. You guys have just done the keynote, SAP and Intel talking together. So talk a little bit more about Intel. You guys have obviously made the bet. I mean everybody's betting on Intel obviously but what does that mean for you and what do you make of this Rhino distribution? What does that all mean? Well I think, I mean HANA itself was a five year really co-innovation project with Intel, the leverage. A lot of what we do is how do you move in pipeline data through the different memory architectures, the different caching levels within the chip sets now and to do that effectively and do extremely, you have to have very low level access to the instruction set and then ultimately to do vector based processing. So you can do huge numbers of integers can be processed in a single clock cycle. You can only do that with a, you can only, you kind of got to pick your horse there. We chose Intel, they've been a phenomenal partner. They're actually taking those same types of techniques that we innovated with HANA with them and they're actually bringing that to Hadoop. And by doing that, so people at the enterprise level are now embracing, saying we see that Hadoop fits in our infrastructure and they want to do that but then when you start encrypting the data you have to have it at rest. Hadoop, which not always the fastest to begin with all of a sudden gets incredibly slow. Intel's work is bringing that back and making that back into acceptable ways but bringing that data so you can have data security and all the different issues you need around data governance into and making that really acceptable but putting the processing power they have behind it. And what we're doing with them is actually bringing the products together so you can have real time when you need it with HANA and the speed and power but then when you have lower value density data that you want in Hadoop or you have just massive other data that you want to archive that you can use the two together but that the IT departments, now that it's being more broadly adopted don't have to know how to write HiveQL, don't if they want to get data out, don't know how to write pig scripts, don't know how to write full tech sentiment analysis, map reduce jobs, down and do that. I can do that between the tool set that they have and we have and that we're integrating over time is to actually make that possible. So generalized IT departments with standard drag and drop type ETL type movements are working and doing advanced work on Hadoop and on HANA making them work together in an integrated function. Absolutely and making this ownable by enterprises that have more broad skill sets within the IT department. Chris, talk about HANA and SAP in context because obviously they don't hype up a lot. Jim Schnabe said to us, said to me and David on a one-on-one last year, you know we like big data but you know it's, we don't want to hype up and SAP is very buttoned up conservative company but his point was, yeah of course but it's really hype market and SAP doesn't usually get a lot of credit as being a big data player when in reality you guys are out there. Can you just talk about, you know some of the big data mojo that SAP has? Yeah, it's funny. We present it and then everyone goes, wait you're one of the only companies we've ever found that actually has every single piece required to do all this. Why don't you talk about it? And we do seem to have- I think we said that on theCUBE too, that's Sapphire. But that's the point. You have a lot of all that. Yeah, so all I've been doing is analyzing interviews and working with them and they all come back with the same thing is and I think that's a function really. We say let's not use the big data. We talk about it and say let's not let's not be perceived of his hype. We want to be given credit for what we've done. So you have HANA at that core allowing you to do and we have even up to, you know, petabyte instances that people can come look at, actually the instances running here in Santa Clara that people can look at that runs a petabyte of raw data and you can query it doing tens of thousands of queries per hour with nothing, you know, a petabyte in less than three seconds. So a majority of queries, worst query of three and a half seconds. I asked Billy Bosworth, the CEO of DataStacks and we have very short on time. We have like 20 seconds to last. I'm going to ask you this for all the noise that's out in the market and there is a lot of noise you walk out here and you see a lot of action going on and it's good actions, just growth in the marketplace but there's a lot of noise. So what do you say to people when they say how should I squint through the noise to look for the signal? What would you advise them? I actually advise them to come talk to us and I mean that by, you're not going to find that written down because we do not do a good job necessarily at some times but a little self-serving. No, of course it is. And I think there's other people to talk to but when it comes to SAP- Talk to us last. Having- Don't talk to us, we're not really, yeah. We don't want your business. We don't want your business. That to administer your problems and the different data storage based on value. Of course you're not going to say don't talk to SAP but no, especially like in looking for use cases like specific data points, like number deployments, what should they look for for evidence within SAP? So when they do come talk to you what are you presenting to them so that you can show them the signal? Is it number deployments? Is it certain architectures? I think what really was my question was looking for more meat on the bone. Like, okay, there's all this noise. Where's the signal? And what can you show them? Yeah, we do. We do a number of deployments. We show different types of use cases that we're now supporting and reference architectures that are actually implemented out there. Solving problems that people have been trying to do. How do I do streaming information off of like gaming systems and do real-time promotions when you die like within a millisecond using statistical analysis? How do you wire that system together? We have those in production and we have those reference architectures and then the references for them to check out. And we post a lot of that on SAPHANA.com as well as, but it's not just about HANA. There's a whole ecosystem of products to support it and we present those out some way. Well, I got to say we were very impressed. Past two years of particular SAP really lays out kind of the gold standard because they had, you know, with the SAP announcement they really laid out this whole in-memory speed of business was a great message. And then last year you guys laid out a lot of use cases. And again, that's what we recommend to folks saying when you want to squint through the noise just look for the meat on the bone look for the sizzle and the steak. So, and you guys are doing a good job. So with that, any teaser for this year's SAP Sapphire that you can talk about without Jonathan and Becker going crazy about that. Without me getting crazy in trouble. Yeah, let's avoid that. I mean, high level. Can you just talk about a little bit of what SAP Sapphire is going to focus on this year? Last year was like, example, we had customers up there, McDermott and Schnabe were awesome and you guys laid out essentially follow-up to the year before which was vision and direction. I think a lot of people have been clamoring around what is this real-time data platform? You have all of these great assets for moving data, for analyzing data both, you know, multi-multipedabyte systems that we have through let's say PSI, PSIQ and speed on ASC for transactional. Well, how are you guys actually going to bring all these together? And we said that's the real-time data platform. And fair enough people said there's not, you haven't really told us how that all comes together. And that's really what's coming out at SAPhire which is here's the things we've delivered, the roadmap, a lot of things that will be delivered that are actually coming out right then at SAPhire that explains exactly how all of those come together along with third-party systems and databases, all in an ecosystem that companies can use to solve all the different data problems we have. And you've got success factors and you've got some business on demand needs to kind of be cobbled together so we're respecting some good news there. Again, we'll see you as a SAPhire SAP, not always getting all the credit for a lot of the sizzle out there but they have the stake. We've seen it at SAPhire, we've been there three years, we'll be there this year. This is theCUBE, I'm John Furrier, we're at SAP, we're at Stratocon for stuff and big data on real-time, in-memory, all that great stuff. So we'll be right back with our next guest at this short break. Thank you.