 Live from Las Vegas, it's theCUBE. Covering Dell Technologies World 2018. Brought to you by Dell EMC and its ecosystem partners. Welcome back to theCUBE, day three in Las Vegas at Dell Technologies World. I'm Lisa Martin with John Troyer. We have been here for three days. There's over 14,000 people here, 30,000 plus more engaging with video content, live stream on demand. We're excited to welcome back to theCUBE, not just back to theCUBE, but back today for a second appearance he's so in demand. Ravi Penteconti, Senior Vice President, servers and systems, product management and marketing at Dell EMC, welcome back. Thank you, Lisa, great to be here. So you have so much energy for day three, but so much excitement, lots of announcements. The theme of this event, Make It Real, is provocative. We've heard a lot of interpretations about what that means for different customers and different industries who are looking to take advantage of emerging technologies, AI, machine learning, deep learning, IoT to make digital transformation real. What's going on in the world of AI and machine learning? Lisa, a lot. Now, having said that, I don't think there's a single industry in any part of the world today that we talk to that's not interested in AI, machine learning, for that matter, deep learning. Why is that so? Just think about the fact that each one of us today is probably creating and generating two and a half times more data than a year ago. It's huge. I mean, when I started out, people used to think megabytes is huge, then it went to terabytes, petabytes, exabytes. Now, I think very soon we're going to talk about zettabytes, right? I'll leave it to you guys to talk about the number of zeros, but sitting that aside, data by itself again. The second thing, when so much of data is being created, data, in my view, has absolutely no value till you create information out of it. Absolutely. And that's where I think companies are becoming more aware of the fact that you need to start getting some information out of it, where it starts the whole engine. First of all, going about collecting all the data. And we have all kinds of data. We've got structured data, unstructured data, and now it's important that we actually get all the disparate data into a format that can now be executed upon. So that's first and foremost what customers are trying to figure out. And then from there comes all the elements, that's the data analytics part, and then you can go into the machine learning and deep learning. So that's the way people are looking at it. And you made an interesting comment, Lisa, which is making it real. This is where people are looking at things beyond the buzzwords, right? It suffices to say, AI is not a new term. We have, I mean, I recall as a kid, we used to talk about AI. But now is when businesses are depending on it to ensure they have the computer to edge. You know, the pendulum swings, right? And 10 years ago, software is eating the world and the cloud is coming and at one point it looked like a future of kind of undifferentiated x86 compute somewhere. Turns out hardware actually matters. And as we, as our applications and data needs have grown, the hardware matters. And so, part of your portfolio is the power edge, set of power edge servers. I mean, how are you approaching that of making the needs of this new generation of software, this massive data parallelism and throughput, real? The great question, John. It's interesting, yes, the pendulum keeps swinging, right? And the beauty is, it's my only hope that as the pendulum swings, we're actually learning too. And we're not making the same thing or the same mistakes. Thankfully, we are not. Now, when people talk about cloud, guess what? To your point, it has to run on something. Software has to run on something. So, obviously, the hardware. Now to keep up with the changing tide and the needs, some of the recent things we have done, as an example with our 840 launch yesterday, you know, NVMe is talking about the town too, talking about some of the new technologies. And customers want us to go out and provide a better way and a faster way for them to get access to the data in a much more faster way, closer to the compute. So that's where the NVMe drives come in. We have got 24 NVMe drives on our 840 today, which is two times more than the closest competitor. Move on to the R940XA. XA stands for Extreme Acceleration. Again, we have never had an XA product. This is the first of its kind that we are bringing out. And the beauty of this is we wanted to make sure there is a one-to-one relationship between the GPU and the CPU. So for every CPU, you have a GPU. It's a one-to-one relationship. If you look at the R940 we introduced earlier, it had, just to give the context to your question, John, it could support four CPUs, but only two GPUs. So we are, think of it this way, we are doubling the number of GPUs. And that's not it. We are actually enabling our customers to add up to eight FPGAs if they want. Nobody else does. And this goes back to, I think Lisa, I think when we started off you talked about FPGAs too. And therein comes the issue, wherein customers don't have the flexibility in most of the cases, in a lot of products out there. We have decided that flexibility has to be given to our customers because they're changing workloads, changing technologies. And even most customers today, they go in thinking that that's all they need, but sooner or later they realize that they need more than what they're planned for. So our goal is to ensure that there is enough of scalability and headroom to enable that to happen. So that's how we as PowerEdge team are building servers today, which actually enables us to provide our customers with an ability to have a headroom. At the same time, give them the flexibility to change, whether it is NVMe drives or any kind of SSD drives, GPUs, FPGAs, so there's all the flexibility built into it along with ease of management. A couple things that you mentioned that I think are really important is that data doesn't have any value unless you're able to extract insights from it. Companies that are transforming digitally well are able to combine and recombine the same data, using it as catalysts across many different applications within a business. That agility is key, that speed is key. How are you, what are some of the things that you're hearing from the 14,000 plus people that I'm sure all lined up to want to talk to you this week about what, for example, PowerEdge is going to enable them to do, you talked about flexibility, you talked about speed, what are some of the real applications that you're hearing feedback-wise from some of these new features that you've announced? Oh, great, so I think again, an excellent question in terms of how the customers are reacting to and what are we doing? So now talking about AI machine learning, think of it this way, right? The permutations and combinations are way too many. And the reason I say that is keeping the hardware aside, when you talk about frameworks that are available today for most of the AI or machine learnings applications, people talk about TensorFlow, people talk about Cafe2, people talk about CNTK. I mean, there's a whole plethora of frameworks. And then there are different neural network methodologies. You hear of DNN, deep neural network. And then you hear things called RNN, there is something called CNN. My point is, there is so many permutations and combinations in the mix that what our customers have come back and told us, going back to where we were earlier, talking about the flexibility in the architecture that we have provided, where we provide seamless scalability on any of the vectors, that they actually love that, we are giving them the flexibility because when there's so many software options with frameworks and every other methodology, we wanted to make sure that we also provided the flexibility, the scalability. And our scalability comes in, whether it is the IO connectivity. We talked about PowerEdge MX, that's going to be coming up soon, that was a preview, but that's why we talked about something called the Kinetic Infrastructure, which essentially enables our customers to go out and run multiple workloads on the same modular infrastructure. Never happened before, right? Or the seamless way we do it now is a lot better than anything else. Likewise, to go back into the R940 XA, we have ability to go out and support hard drives, SSDs, FPGAs, GPUs. So the feedback has been that our customers are really excited about the fact that we're giving them the flexibility and agility to go out and match to the needs of the different workloads and the different options they have. So they love it. I talked with some of your team yesterday and I was really impressed as they talked about the product development cycle. They said that we start with the customers and we start with applications. And then we figure out what technologies are now appropriate to build and what combinations. They don't just start from like, let's throw the newest thing in because we can. As you talk to ICIOs and enterprise architects, it used to be, you just do a server refresh and kind of just check the box and push the button. Now you've got to look at cloud readiness and what I keep on-prem and what I keep off-prem and which, what's going to fit in my applications? What are you hearing from customers and how are you trying to educate them on how to approach their next refresh? Well, I think even refresh is probably a bad frame. Their next set of applications that they're going to have to build in this digital transformation. You know, John, this is actually no different. I mean, let's step aside from the compute world for a minute. Let's pick up an automobile industry, right? If you get into the automobile industry, a family might say they need a sedan or a family of five or six with young kids might say they want a minivan, right? And maybe now the kids have grown up or you're still in your 20s or 30s and some of the folks would love to have a sports car like the McLaren that was up on the stage. I'll take that one. Jeff, I know, I would love that too, right? So my point is, when people are trying to decide on what is it they really want to buy, they actually know what they're looking for, right? A family of four doesn't go in and say, I need a two-seat car, for example. It's a similar thing here as people start looking at the workloads first. They come in and start looking at mapping. Hey, this is the kind of a workload we have now. Now let's start looking at what infrastructure can we provide behind it? You know, and even if you look at our, something that we have announced in the past with the 740 XD. So we have a 740 version and a 740 XD version. XD there stands for extreme density. So if customers want a two CPU box, a two-U box server, but they want more storage, then they have XD version. But they decide that storage is not really crucial. They just need the compute, then we provide the 740 on its own, the R740. So my point being that accentuating the point you raised is it's always nice to look at the application, look at what its needs are, whether it is memory, whether it's storage, whether it is the GPUs or CPUs, and then look at how it transposes itself over the next few years, because you really don't want to acquire something and then really decide later that you've run out of room. It's like buying a home and then you know you're going to have your kids, you're going to raise a family. You don't probably want to start off with a single bedroom and you know you're going to have a family in a couple of years. My point again being that, that is where the planning becomes absolutely important. So we are planning and the planning phase is crucial because once you have that right, you now can rest at ease for the next few years. And as we do that, one of the other fundamental design principles of PowerEdge is that we want to really support the platforms for multiple generations. Case in point, when we came out with our PowerEdge M1000E, we said that we will guarantee support for three generations of processors. We actually are up to the, we're up to the fifth generation as we speak right now. And our customers love it because nobody really wants to go ahead and buy more servers every few years if they can go back with their investment they've made and you know ensure that there is a room to grow. So to your point, absolutely the right spot to start is start looking at the workload, start looking once you have pegged it, then start looking at where you're growing and what your needs could be and then start connecting the dots, nothing you know you would be coming up with a better outcome for the long run. We had the opportunity to talk, John and I just an hour or two ago with the CIO with Boss Iyer and one of the things that was interesting is we talked to him about how the role of the CIO is changing to be really part of corporate strategy and business strategy. As you talk with customers about building this infrastructure to set them up for the flexibility and the agility that they need, allowing them to make the right decisions for what they need, but also scale it over time. How much are you seeing the boots on the street that you're talking to have to sell this up the stack as this is fundamental to transforming IT which is fundamental to transforming our business into a digital business? Very true. By the way, Bask is a great friend and a collaborator. We certainly look to you as I say, the saying goes, you're here on Darkfoot. So we work with Bask and team very closely because as a CIO of a large corporation himself, I mean, we learn a lot. There's nothing better than trying to walk in the shoes of our customers. So going back to the comment you made, Lisa, is most of the, by the way, most of the customers today, the CIOs who are now becoming not cost centers, they're becoming profit centers. That's what Michael D'Oceano on Monday is. Absolutely, and he's absolutely right. Michael is absolutely right because most of the organizations we speak to today, on an average, I would think that the number of CIOs we talk to has probably been dialed up because we see the kind of questions that are being asked of, right? To the point we were making earlier, they're not looking at making point purchases for something that would satisfy them for the next 12 months or 18 months. They're looking at the next horizon. They're looking at a long-term strategy and then they're looking back at the ROI. So what is it I'm able to go back in and provide to my customers internally, whether it is in terms of the number of users or the performance, whatever the SLAs, the service level agreements may be internally, that's what they're looking for. So towards that end, the whole concept of ROI and TCO, the total cost of ownership and the return of investment nowadays is probably a much bigger talking point that we need to support with the right factoids. And I think that's becoming crucial. And the CIOs are getting more engaged in those discussions. They're never in the past. And so it's just not about feeds and speeds, which I guess anyone can look at spec sheets, not as exciting, but there are things beyond that that I think are getting more crucial. Well, Bask said, drinking your own champagne, eating your own dog food. I like champagne and dogs, I'll go with both. Why not? I just saw... We've got the therapy dogs at the store. Therapy dogs, exactly. They're fantastic. They're great, they're great. So last question in the last 30 seconds or so, biggest event, 14,000 as I said, expected live over the last three days and tens of thousands more engaging. Any one thing really stand out to you at this inaugural Dell Technologies world? The most important thing that has stuck for me is that human progress is indeed possible through technology. And this is the best showcase possible. And when you can enable human progress, which cuts across boundaries of nationality and boundaries of any other kind, I think we are in the winning streak. Well said. Ravi, thanks so much for coming back today, couple times and hanging out with us on theCUBE and sharing some of the insights that you're seeing and that you're enabling your customers to achieve. Thank you, Lisa. Thank you, John. It's been awesome. It's always wonderful being with you guys, so thank you. We want to thank you for watching theCUBE again, Lisa Martin with John Troyer Live, day three of Dell Technologies World. Stick around, we'll be right back after a short break.