 Well hello there, Mark Risen Hopkins here, coming at you once again from Google IO 2013. I'm here with the guys from MapR. We're going to talk about a lot of interesting things that from both this event and last year's event, because last year's event was so awesome, we're still talking about it. And some new things that are kind of in the cloud war space, broadly about Google Compute Engine, which they are a participant in as well. So let's just start off, last year at Google IO, this event, you did this amazing demo in which you demonstrated the difference in cost between deploying on Gridiron or for, I'm sorry, I should say for a MapReduce operation on doing on bare metal versus using Google Compute Engine. And the difference in cost was amazing, very stunning, made lots of headlines. So let's refresh people's memory on that, and then we'll move on to what's coming up this year. Yeah, so one of the things that we showed last year at Google IO was actually our ability to leverage Google Compute Engine in Google's cloud. What we were able to demonstrate last year was the ability to sort a terabyte worth of data and how fast you can sort one terabyte worth of data. The numbers were astonishing. We were able to break the existing world record that was actually there. We were able to crush that record by doing it, I think with a third less cores, half the CPUs that are actually needed. It was an amazing achievement number for us. And really, when you looked at the cost, I think it was like three to $6 million or what it cost you to do it on real equipment. And I think we got it down below $20 to go ahead and do that processing. And so it was pretty amazing. It was a really big buzz as we were really launching and kind of getting computing out out in limited preview and getting customers on it. So in this year, you're kind of demonstrating some stuff with a client of yours named RAP. Very interesting, you've got the demo running behind you, a real-time live demo. So talk about what you're doing there or what you and RAP are doing there. It's very interesting, it's actually very cool to kind of see the little tag clouds form in real-time there. So what are we looking at here? So really what you're taking a look at is real-time data or stream processing of the data. We were looking at it for the MacBook timed out there. We'll talk about it for just a second while we wait for the computer to come back. We're doing real-time processing of the data. So taking real-time from actually a gaming application and what it's doing is it's pulling in user information. What are the actual users doing across the data? Who are the most active users on the game? And what it's pairing it to is what the actual user is doing. So it's an actual online game where there's, people are making moves, getting weapons, et cetera. And what you're able to take from this data is, as a developer, are my seeing errors? Am I seeing lots of logouts of a user? Maybe I have an issue with my gaming application or maybe there's a monetization play. Maybe I want to offer them a new weapon or something that they use as part of the game to kind of entice them to actually use something. So be able to find gaming pain points in real-time. Pain points in real-time and what they're actually doing. So it's a real-time demo as we're really starting to see this shift from a Hadoop standpoint as Hadoop's really kind of been regulated to more batch-based processing. It's now moving more towards real-time and stream-based analysis. And this is one of the demos that we've done to kind of show what we're doing there. Right. So it's very interesting. Maybe we'll get some B-roll here in a second of that and splice it in. But I also want to talk about, and this is a big topic right now kind of in the blogosphere. Everyone likes to use the word cloud wars. And you guys have a foot in a lot of different worlds, different clouds, obviously with Google Compute Engine. Of course you work with the other clouds as well. So talk a little bit about your perspective because first of all, is it a war? Is it not? Is it a war? Is it just new supercomputers coming online at different speeds? Yeah, I think it's really more of a second. I'm not sure if it's a war and really as you look at it, the space is huge. It's really, in some respects, really green-field in a lot of cases. So I'm not sure if it's really kind of hand-to-hand battle. It's really kind of out there. Who can get in front of the most customers fast enough? There's some customers that are going to prefer Google's cloud that are using storage, BigQuery, et cetera. And they're going to prefer other clouds. And I think it's really, for the cloud providers that can provide the most dynamic services and the most sticky services to their end users are going to want to capture the customers. And I'm not sure if you're going to be able to find a de facto winner and loser, but I think really who wins in this is actually the consumer or the business who's actually getting these services. Because as more and more competitors come into the space and as more and more providers, the cost of those services actually come down over time. So it's only going to be a benefit to both the consumer and the business. So Wikibon Analyst, Chief Analyst Dave Vellante, likes to use the phrase, horses for courses to kind of describe a lot of different things in storage and enterprise. And I think what I'm personally gathering, and maybe you can tell me if you agree or disagree from talking to different partners within the Google ecosystem, is that there are certain places where like Amazon is far ahead and shoulders above where Google Compute Engine is. Specifically, I like in areas of maybe like video processing in the cloud or kind of specialized tasks where they've just had Amazon's got a lead, right? They've been around doing it longer and offering it as a service. And then there's other places where Google has focused on hyperscale for web application, mapping MapReduce, obviously a pioneer in that space too. So I guess the question I have for you would be, so how do you feel BigQuery stacks up against the competition in the other hyperscale players? Is it a head and shoulders winner? Is it a comparable like feature set for feature set? Where is it in kind of the hierarchy or the ecosystem? You know, I think it's more of an in-user question. I mean, as a partner, we understand how to interact with BigQuery, but I don't think we see it from that standpoint. So I'm not sure if we're probably the best to comment on kind of whether BigQuery's heads and tails above the rest, but I think one of the things that we've seen with our customers that are actually using us is, having that in-to-in data pipeline really matters. So being able to take services from like Google App Engine, being able to do log file analysis, leveraging MapReduce to go ahead and do that, and then have that being processed from BigQuery, I think makes a lot of sense. And I think in other clouds that have similar type services, providing those high-end services or enriched services to your end users really matters. So I don't know if one's a de facto leader or one's heads and tails above the rest. I think some of it also depends on the customer use case and what they're using and where their data is. One of the things is part of it is your data lives in the cloud, and so some of it could just be where your data lives at the moment and who's providing you the services that you need. And who's your neighbors, right? Which cloud is your neighbor's data in? So last question, because we're running out of time, but I do want to ask, I've got this friend of mine. I go to every once in a while. He works for a very large bank that many of us use. I won't say the name because he doesn't like it when I do that on air. But I talk to him, because he's getting into Hadoop. They've been dabbling with big data for a number of years, but being a large bank, they've got, of course, restrictions and red tape internally. But I showed him the announcement from last year. Look at how much money, because they're building their own, like their own Hadoop servers in-house. Look at this thing, you can save millions of bucks. But it comes down to security, security, security. And being that the public cloud is a nascent field, how do you see the progression for financial sector? Is that their adoption is obviously going to be much slower than perhaps other sectors of the enterprise? Where are we in that? Is that particular very large bank alone in that, or are the other banks kind of coming along more quickly, or is the whole sector going to lag? Well, I think overall security needs to be addressed. And I think that's one of the things that kind of goes back on the shoulders for the big cloud providers, be it Google or any other cloud providers, security's got to be addressed. And especially when you look at financial services or healthcare organizations with HIPAA and PII information, I think those are definitely going to be addressed. And I think there's going to be two ways in which those are addressed. One is they're going to specialize clouds, maybe for financial services, like FISR of clouds, et cetera, same way that some of the cloud providers do for the government, or they may actually have to just put in strict regulations across all of their cloud-based properties and say, come on, come on, but it's going to be super secure. So I really kind of see that playing in strongly with the cloud providers. I think it's a question of kind of where we are in that adoption curve. For those folks, I still think we're early, but for a lot of these new technologies like Hadoop, we're still early in that process, but we're starting to see a lot of interest because having large amounts of data and to be able to have unlimited, or seemingly unlimited processing power to process that data at fractions of the cost, I think could be huge over time. The power is clear. And for my friend and his organization, the hurdle is simply just getting past these 1990s-style bureaucratic regulations about what you can use and what it's considered secure. So I would tend to agree from my experience with your assessment. It's just one of those things that's going to have to come along. So yeah. So yeah, I appreciate your time. There you go. I appreciate your time. And we're going to be bringing you more and more coverage from the floor of Google I-O 2013. Stay with us, and we'll be back soon.