 Live from Las Vegas, it's theCUBE covering HPE Discover 2017 brought to you by Hewlett Packard Enterprise. Okay, welcome back everyone. We are here live in Las Vegas for theCUBE's exclusive three-day coverage of HPE Discover 2017. Our seven years of covering HPE Discover, two years into the new HPE Enterprise Discover 2017. It's our show. I'm John Burrier. I'm my co-host Dave Vellante. We have a multi-year seven straight, I think seven straight years, Ray D. Byer, Vice President General Manager of HPE Mission Critical Solutions. Welcome back again. Oh, thanks for having me. I absolutely love being on this event and talking about all the cool things that are happening in the technology world and having and hearing from our customers what they're doing with. Well, you remember Randy, we're in a little table and you know, the show floor. Yeah, we've gone kind of Hollywood here as opposed to the early days. And Mission Critical even more, even more with HPE and all this great stuff going on. Amazing. Our other guest, Dr. Stephen Pratt, Corporate Technology, our center point energy. Welcome to the queue. You're also a CUBE alumni. Welcome back. Thank you very much. So guys, what's going on? I mean, we're seeing Mission Critical workloads. Obviously staying on prem. Wikibon just released their new true private cloud report, which is, hey, cloud's here, of course. But still on prem games here and still going to be all those Mission Critical workloads. They're not going anywhere soon. Yes, some dabbling in the cloud and some SaaS happening. Well, it turns out, and I'm going to let Dr. Pratt really talk about what's going on at the customer level. The world's hybrid. There's some things that'll move to the cloud, but when you start to look at the Mission Critical applications, they're processing payments. They have social security numbers in them. They have healthcare information in them. They're managing networks that if somebody hacked into them, you could take down a banking network. You could take down a telecom provider. You could take down a utility and all of these sorts of things. Security, there's a lot of reasons people want to have their control or their applications and their data, and that's not changing. But what hasn't changed is they need to run all the time. They need to be able to do that in an environment that's going to give them flexibility. And the other thing that we're seeing that's just fundamental is this confluence of the systems that you use to take action. How do you schedule somebody? How do you move a spare part somewhere? How do you take action on a system or a transaction? And then you take all this IoT data that's coming at it. Everything has a sensor in it. I would bet this headset has a sensor in it, knowing you guys. But what do you do with that? How do I take all of that sensor data and marry it to the transaction system so I can in real time right now do something with it? And that's what CenterPoint Energy's really doing. It's just about CenterPoint what you guys are doing and how you fit into this equation. So CenterPoint Energy's an energy delivery company. And when we talk about hybrid cloud or whether you're off-prem or on-prem, cloud solutions, those types of things, it's really not dictated by us and making a choice. It's really determined by your customers and their expectation and what they're wanting to see in terms of services from whatever company they deal with. In our case, it's energy delivery services. We have 2.3 million customers at CenterPoint Energy. All of them now have sensors on the side of their homes. They're residential customers. They are smart meters. They are meters that can tell us things. They can tell us their state of health, those types of things. They generate data. That data needs to come back and be assessed. It needs to be determined as to whether it is something that we need to address, those types of things. So the whole cloud, the off-prem cloud, the use of IoT, the intelligence of things, that becomes a big part of meeting our customer's expectations. Customer really wants two things. They want uninterrupted delivery of energy services and they want to make sure it's done in a safe way. The fact of the matter is, they'd rather forget their utility, their provider of energy services because if they're calling us, it typically means they have an issue that we need to address. It's not like what HP is trying to make things invisible. That's what the customer wants, but they want the energy. They don't want to lose that power, right? It's like, why buy? Why don't you have it? You can't go back. All right, so you're in the middle of the IOT. You're living the IOT dream, as we would say, depending on how you look at it. But take us through that because now you are preexisting conditions with that IOT. You got the meters, they have sensors everywhere, bringing that into an IT environment. Take us through that journey and how that relates to the things that you got to redo or shift, how your IT consumption is changing. How does that impact you? Absolutely, it's a completely different model. If we just think about data itself and the data that's generated by these IOT devices, these meters, if you will, 10 years ago, we had about 80,000 meter reads we did a day and it took 400 people to do that. Today, we do meter reads 221 million times a day and it takes no people to do that. That's really what IOT devices are really intended to do. They do the work for us. It allows our people to focus on other things. And so we have changed from what is a traditional utility to a digital utility, which means we are driven by technology. We're driven by data. We're driven by technology that allows us to consume data and assess that data and use that data in some cases in near real time. You need speed. You've got to have the performance. We absolutely have to have the speed of performance. At the heart of this today, our consolidated data platform has become an SAP platform which is really driven by Hewlett Packard Enterprise Technologies. It's a very, very high performance platform. It's a very, very scalable platform. It can handle all the data we throw at it. It can assess that data in real time. But also, if not in real time, we can use that data historically for other purposes as well. So we've really transitioned from a company that was very manual in how it delivered energy to a company that's very automated in the way that we deliver energy. And so as a result, we're better able to meet those customer expectations. And those meters are obviously they're connected. They're instrumented. Can you help us understand the data flow and how that all works and what you do with the data, where it lives? So every meter actually has a microprocessor in it. Every meter has the ability to process some level of data. Every meter has the ability to send you information, event information about what's occurring at that residence. The flow of the energy flew that meter, if you will. It starts with the meter. That ties back into what's called a cell relay, which is another intelligent vice. From that cell relay, that ties back into what's called a data collection engine, which is an engine that really aggregates data that comes from a population of meters, typically four or 500 meters, have one collection engine associated with it. From that data collection engine, it then flows back through our network and ultimately into our data center. And every level along the way, we can do some level of analysis about that data. So what that does is allow us to not move data that we don't necessarily need to move or not manage data that we don't really want to manage unless it provides us some additional value in doing so. You have multiple data collection points. Absolutely. And then ultimately the data that needs to end up in some whatever data lake or whatever it is ends up there. Exactly. It's enormous amounts of data, a little bit amount of data, medium-sized data. What are we talking about here? So every meter and typically every sensor doesn't generate a lot of data. The real challenge is all of those meters generate data in real time. We have to, as a regulated entity, maintain that data for 10 years regardless of whether we use it or not. So on any given day, 221 million events being read and stored doesn't really amount to a lot. But after you look at the accumulation each and every day, each and every year, for 10 years it becomes a massive amount of data. So the real benefit is being able to leverage that data to create additional value because you're going to have to manage it regardless. And that lands in the, is it a data lake somewhere or John hates that term, data ocean? I hate data lake, more like an ocean. So we actually don't have a data lake and we don't refer to it as a data lake. We refer to it as an advanced data management platform. And in this particular case, Achilla Packard Enterprises underlying hardware with SAP's HANA on top of it. If you want to call that a data lake, call it a data lake. But it also gives us a flexibility to process data in real time in some cases but also process data in a longer periods of time so you don't have to consume all of that power all of the time. But it's a data in an analytics engine which is essentially your data warehouse, is it not? Absolutely correct. It's not really a data warehouse because it's got embedded intelligence but it has a data warehouse built into that service. I mean, being a techie though, I get a little bit parochial on these definitions too because data lake oversimplifies things. It makes it sound like, oh, just going to throw it in there and that's where the data swamp term comes in because it ends up sitting there and no one knows what to do with it when you bring intelligence and you need to make the data work for you, right? This is the key thing. Yeah, and you start to look at this and across all these different industries, it's the fact that you don't just put it in there but you're right, it goes into swamp and we never look at it and yes, you're regulated so you have to keep it but it's what you can do with it. You think about an energy company, there's seasonal patterns, right? There's daily patterns, there's probably holiday patterns or there's weather related patterns that drive you in things. You don't see those until you get some longitudinal data across it. So in some cases you can take action immediately on something you see. I'm guessing neighborhood browns out. You know it instantly now, right? Or gets even a dip in voltage. You can go do something with that. Let me get to the engagement level with the customer and actually, I mean people these days are very conscious on energy studies. So I think, you know, having data and saying, hey, if you just do things a little bit differently, maybe we have some bitten sentence. So I think this would enable business model transformation. It absolutely does. We're not a competitive retailer. We don't sell energy, we move energy but we provide information that we gather from the meter to the competitive retailers and they use that in a variety of ways. In some cases, new services that they can deliver to their customers and they do. We run a number of analytic scenarios but the intent is to be able to ultimately route around an outage or inhibit a potential outage by knowing as much as you can and as quickly as you can know that. And that's the real power of having an analytics engine and that amount of data. And you also enable some guys to just say, hey, you know, let's enable an API to our clients and then you get some things, things just kind of magically happen when you have that kind of capability. It's absolutely right. It really is now a data-driven, intelligence-driven organization and it's an organization that really is from an operations perspective, it is dependent on that data. It's dependent on the amount of that data and it's dependent on that data being able to be consumed and utilized in a valuable way really almost in real time. Well, and it's funny, you talk about that being able to reroute the energy grid, major European supermarket chain has a very similar analytics platform running almost identical configuration to yours. They use it where they say, wow, the store sales in area X, this product's flying off the shelves but an area Y, I've got too many of them, they reroute trucks on the fly that says, stop delivering them to the area where they're not selling, reroute them to the area where we need more of them. This is rerouting the grid if you're a retailer and making sure the product's in the right place. Hey, I just can't wait for blockchain to be into this conversation. We're having a good time right now. Throw some blockchain in there and we'd hit like all the action. Actually, I can throw some blockchain in if you like. So we are regulated, so we have a market and we have an energy regulatory commission of Texas that we have to send transactions to today. These could be digital transactions. It could be based on a digital ledger. So blockchain is not without outside the realm of possibilities for a way that we would actually move transactions back and forth within our Texas market. So blockchain could ultimately figure into this. Randy, we're going to include you on our cube coins. I see how we're starting. Day out and look out for cube coins coming to a blockchain near you. Of course, great, we love social, we love digital guys, congratulations. Randy, give you the final word, mission critical. As we heard from Alon earlier, HPC, you're seeing AI, you're seeing a lot of where compute is happening, machine learning, really the data solution is coming. This is your wheelhouse. What's your, what's going on here to discover and what's the big news? I mean, so you're watching all sorts of different things happening. We're innovating in this in-memory scale-up space. I've got some of the biggest things. Data is coming close to in-memory where you can compute right on the fly. We're innovating from that standpoint. I've got massive, I can take you in my labs and show you 48 terabytes of main memory all flying at the same problem. I mean, this is just unreal. Pro compute at things is what's going to be like, lighting a bolt of compute, any problem. But the other part of it is, if I look at my software, I've got the nonstop software stack, right? My, the runs all the payments around the world. You talk about blockchain, we just ported our three quarter to nonstop. So you can now run blockchain on nonstop. And it all runs in an open stack cloud if you want it that way. Totally virtualized, totally software. So we're all about giving people flexibility. You want massive scale-up memory to solve the kind of problems that CenterPoint does. I've got that. You want systems that run without fail, that you can run forever. I can go manage that and I can give it to you in a private cloud, on-prem, I don't care how you do it. Mission critical and high performance computing are becoming just fundamental. They're high growth areas because people expect things to work all the time. And you start to think about this. It just gets embedded in the fabric of what you do, right? And if you can make it easy so that the, your customer doesn't have to do all the R&D. Absolutely. That's Dave's been pointing, Wikibon research has been pointing out that, that non-differentiated labor and R&D that you're going into can be abstracted away. If you guys can provide that, then the customers can spend all their time over now what they need to do. We engineer solutions to these problems. We talk about what are you trying to do? You're trying to move energy. You're trying to manage a supermarket. You're trying to go deliver things in a real time where you can give a customer a reason to visit your store. We talk about those workloads. Describe the business process to me. My job is to say, I'm going to pull all the technologies. And by the way, it's compute, it's networking, it's storage, it's software innovation. It's the know how to package that in a way that to be honest, for most customers, what they really want to say, how do I get that level of experience in terms of building an advanced data platform or I could take it off the dock and plug in a power cable and a network cable and that's what I have to know. I want to know my business. You guys know the technology. Great staff. Randy Meyer, Dr. Steven Pratt. Thanks for sharing the insight. Congratulations. You are living the IoT dream. We are here at the edge of the network with theCUBE at the HP Discover in Las Vegas. Bringing you all the action. I'm John Furrier. Stay with us for more day three coverage of HP Discover after this short break. Stay with us.