 Live from Las Vegas, Nevada, it's theCUBE, covering EMC World 2015. Brought to you by EMC, Brocade, and VCE. Hi buddy, it's Dave Vellante back with John Furrier. This is theCUBE, we're live at EMC World 2015. Anthony Smith is here, he's a converged infrastructure specialist at Lotus Formula One Team. Anthony, welcome to theCUBE, it's great to see you. Thank you very much. So, how is it that you guys go so fast? If you saw the keynote Monday with the motorcyclist, did you see that? I'm not going to say anything like that. It was pretty funny, I tweeted that out. Check my Twitter stream. So, welcome, cool stuff that you guys are doing. And you're helping educate the world on how data is so important to you guys winning races. Absolutely, yeah, absolutely. We're looking for hundreds of a second on the car. Everybody out there, we're competing with some very, very clever people. So, everybody's doing similar things, and we're trying to beat them. It's a competitive advantage, if we can get a competitive advantage through our data, that's the way we win races. And it's a real horses for courses type of thing, right? I mean, in terms of every course is different, your strategy changes, some courses you're worried about tire wear, other courses you're not so much worried about that. How does the data affect those outcomes? Oh, we're constantly gathering the data from the cars. It's going around, we do tests, and then join the race as well. There's over 150 sensors on the car. We get about 60 gigabytes of data per race weekend, and we're using that to analyze the performance, analyze our strategy, make our decisions, and we're doing that live. So, there's a constant stream off the car, about two megabits. And then we use that, and we use modeling. We do proper statistical modeling. We do simulations. We simulate about 20,000 different race outcomes every lap. So, what are the examples? Every lap. So, we're analyzing what everybody else is doing, but everybody does the same thing. We're looking at how fast they're going, we're looking at, we make estimators to their fuel load, their tires, we try and guess, we've got to try and second guess them, and we've got to try and beat them. It used to be art, John. So, give an example of where you guys were changed your strategy, or avoided catastrophe, or won, or give some practical examples for the audience out there, it's a lot of sensors, a lot of data. I mean, all the time we're making decisions that affect the car. We've got to get the car to the finish for one thing, and so, that affects our fuel load, our fuel strategy. So, typical race, say, one kilo of fuel, so two pounds of fuel in the car is going to be worth about 300th of a second lap, which doesn't sound much. But, over 70 laps, that's two seconds. And then, we've had races where the race has been won by 0.7 of a second. You know, we had a race last year where one kilo of fuel would have been the difference between second place and fifth place. So, that's just one kilo, and that's just one part of the car. There's so much going on. There's the tire wear, there's the brake wear. We can change the fuel consumption of the car as we're going around the track. Okay, so let's go onto the hood, so to speak. With EMC, what are you guys doing with them? From a tech perspective, how do you roll it out? What's going on? Well, EMC are a technical partner of Lotus Formula One team. So, they supply us with all of our back end systems, effectively. So, we've got a couple of V-blocks at the factory. So, that gives us, we've got two data centers. So, we've got an active active pair. So, we're completely resilient there. That's for all of our data and our analysis there. And then, at the track, we take a smaller V-block, a 220. We take that to every race that we do. So, that's- The mobile V-block. Mobile V-block, yep. And that's 20 races, and that flies around the world. And it sits in the garage, pretty much next to the car. With no cooling, nothing like that. Not at a data center, you know. It gets a hard life. But we rely on that, and it's critical to what we do. And what did you guys show at the show here? What are you guys doing in the booth? Well, we've got, over at the booth, we've got a real race car, and we're doing wheel changes. So, there's a competition going on, just to show people how difficult it is. Because people see it on the TV, and we can do it in 2.1 seconds. Change all four wheels on the car. Wow. That's pretty good. So, we're giving people a bit of an experience of that. To give them a bit of an experience of Formula One, and what we're actually doing. We've got simulators. We've got these fitness response machines. Because our drivers have got to be super fit, and super agile. Would you say your adoption of this sort of data orientation occurred in a big spike, or has it been sort of a slow maturity climb? Can you describe that? It's both, you've got a gradual sort of gain, but also you get these big jumps in technology, which is things like private cloud virtualization, that sort of thing. Because when we go track side, we're running about 120 servers. And so, that's running on our B block. We can't do that with physical servers. There's no way. We pay for the freight that we send around. We pay for every kilo of weight we have to send, we have to pay for. So, being able to virtualize it was a massive step, and most of the teams are now doing that. So, it is the most competitive business that you can think of. You're talking about hundreds of a second, make a difference. And you're all chasing the same model, the same price. People leave companies, they go to different teams. How do you maintain a competitive advantage? We could do everything we can. We've got some very clever people. We're trying to think of every possible way that we can get that advantage, whether it's do parts on the car, whether it's do data manipulation, data analysis, simulation, all of that sort of thing is all key to what we're doing. And that's why we work so closely with EMC in terms of having an agile system that we can do changes and that evolves and that improves. Imagine trying to minimize the amount of what I call non-differentiated heavy lifting that you're doing. Because you don't want people fiddling with patching, trying to innovate up the curve. You've got converged infrastructure in your title. That's a relatively new title. It is a very new title, for me, because I haven't been with a company that long. But even the industry is new, right? Yeah, I cover all the, basically, all the EMC real estate that we have. So it is new, it's evolving. As they say, that job wouldn't have been there, sort of, converging infrastructure specialist sort of a couple of years ago. There wasn't, and we're moving, we got converged and then we're going hyper-converged. So, is there going to be a hyper-converging for such a specialist? Well, it's interesting, right? The hyper-converged piece, it's the software-driven. How does that change? What are your thoughts on the so-called hyper-converged? Well, of course, we're looking at it. It's progress. If we can minimize the equipment that we take, we can up the reliability, we can up the flexibility that we've got, the ability to do more with it. We only take one person to do the IT track side. That's it. So the kit has got to be bulletproof. Because it's got to run. We haven't got the time to fix it or anything to go wrong. It's just got to work. What's your experience been with that? I mean, sometimes things go wrong on the track. I've had a lot of experience in Formula One and I've had a lot of things go wrong, but the stuff that we've got with the V-Box has been great. I can say that, you know. So your IT infrastructure has not been, you know, like some of the problems you have with the mechanical issues or engine issues. I had a terrible race about eight years ago. I blew up 15 UPSs in one weekend. I don't know where I got them from. So what would the world be like without VCE and the V-Block? You don't have stack servers. What would the alternative be? Well, you know, we moved from, when we joined up with AMC and VCE, we moved from a completely disparate system. It was storage from one. It was compute from another. It was all over the place. But with the VCE stuff, we've got a completely convert system. It's one contact. It's one system that all works together. We do the matrix upgrades. We know everything's going to work. It just makes things that much easier for us. So yeah, then I got to ask you, when you're having a pine sitting around with the guys and the gals and you're, you know, getting creative, what do you think about the future of what you do next? I mean, because you can instrument the hell out of everything, right? So where is the creative juices going next? On our side, on the IT side, we're shrinking the footprint that we've got as everybody is. We're reducing the power consumption. We're doing all the normal things that people are doing in data sensors, but on a smaller scale. We're also moving towards the proper meaning of big data and actually going right across our data sets to actually get more out of it. Because, you know, a race engineer can only look at a certain amount of data. He's a pizza person. Whereas if we can automate that, if we can get proper big data systems in there to actually look across and through those data sets, we think there's advantages to having that, for sure. When you look at EMC, the federation, there are essentially three different companies, actually four, if you look at RSAs in there too. Are you tapping into other parts of the federation? Or is there a wish list we have? We've got a wish list where we're trying to get, you know, there's parts of the federation that we've got, we're covering, you know, we've got the V blocks, that are perfect for us, you know, we've got the whole of the big data side of it, we've got the security side of it, we've got the VMware side of it. All of those parts fit in with what we're doing. So it's absolutely, you know, it's great to be involved. What kind of future requirements do you need from EMC? What do you want them to do to make your life better, to make you more competitive, help you win more races? Well, we're off to anything that we can get that'll give us an advantage. So whether that's raw compute, whether it's storage, whether it's flexibility. So we're looking at, you know, we're looking at the new stuff, we're looking at V specs blues, we're looking at VX racks, we're looking at the big data side of things. We are fully sort of EMC throughout. We've got tiered storage and the archives with our apnoces and the data domains, all that sort of stuff. But we just wanted to push it further and push it harder to work better for us. And a lot of the data analysis that you're doing, of course, is real time. What about after the fact? What are you doing with that? Well, we keep it because of course we need it. We're then analyzing it, we're feeding it back into our systems. We're looking at different ways, comparing it with previous years. We're feeding it into our simulators that we've got in the UK. We've got R&D facilities so we can play back a race once we've done it or even joined the tests. We can play the data back to a car that's on a, what we call a seven post rig. So we can drive each of the wheels and it's as if the car's going around the circuit that we can refine the performance and get it better for the next race because we know that we've got to move on because all the other teams are moving on doing the same sort of stuff. So when we go to a race, we don't just go with a car and then set it up. We go knowing how the car needs to be set up and then we just make fine tweaks to it because there's so much that can be changed on the car. I think there's 1600 separate parameters that we can tweak on the car that we can change. And they all need to be right when we get to the circuit because we only have an hour and a half, two one and a half hour sessions on a Friday to basically set the car up. We don't test during the year, we've got two tests during the year, that's it. Three at the beginning of the year and then two during the season. So we've got to simulate. We've got to take that data and feed it back into the system and then progress. How much data are we talking about here? It's about 60 gigabytes per weekend that we collect off the car. We've got, I'd say we've got about 150 sensors. About a two megabit stream off each car. 25 megabytes per lap, that sort of amount of data. So it's not huge. We're not talking about the petabytes. We generate plenty back at the factory as well with our computational modeling. We're producing about, I think it's a terabyte a week. Okay, in terms of the sort of backend data analysis, what kind of techniques are you using to perform that? We've got a lot of in-house tools that we use at the moment for analyzing our data. So, and of course we were finding those. We've got our own software developers. So it's all, it fits in with the whole formula one thing which is very, very iterative. We do something, we refine it, we do it again, we get better and we get better and we get better. And so the communications from the onsite data center to somebody in the pit crew to the driver, what is that individual that your systems are communicating to, what does he have in his hand? Is it a mobile device? Has he got a little laptop there? Is he going to, you got a mainframe? Well, of course we virtualize the servers that they sit on the VBlock. People have laptops, they're there analyzing the data by using the powerful compute capacity that we've got on the VMs to do the data crunching. And they happen, they've got statistics packages that we're running all the time as we're gathering the data. So we can see how the car's behaving and then we'll go back to the driver, say during a test session, and we'll show the driver what he's doing and where we think we can make improvements. And we say, right, at this point you need to be doing something else. And we will show him what's called an overlay where you put two graphs on top of each other. But it's not quite that simple but that's one of the things we do. In the communications to the driver, is it all sort of audio or does he have a dashboard as well? No, we can only, we can talk to the driver. Yeah. That's two way and we get the data from the car but we can't transmit back. Part of the rules say we can't transmit back to the car. And he's fairly busy in the car. That safety issue, right? Yeah, I mean, yeah. He's fairly busy in that car. So he gets his lap time, he gets information to say whether he's faster or slower than he was, you know, than his fastest lap, that sort of thing. And we're giving him all that feedback all the time which comes, of course, from the data. It comes from our strategy decisions which come from the data. Interesting. All right, Anthony. Listen, thanks very much for coming on theCUBE. Really appreciate your time. It's a really interesting story. Thank you. I appreciate it. Keep right there, everybody. John and I will be back with our next guest right after this. This is theCUBE. We're live from EMC World 2015. We'll be right back.