 Live from Las Vegas, it's theCUBE, covering Dell Technologies World 2018. Brought to you by Dell EMC and its ecosystem partners. Well, welcome back. Glad to have you live here on theCUBE as we continue our coverage of Dell Technologies World 2018. We are live in Las Vegas. We're in the Sands Exposition Center. I'm with Keith Townsend, who had a heck of a night last night. Just a good chicken and waffle Las Vegas night. You know what? One o'clock in the morning, there's chicken and waffles here in the Grand Luxe in a viewable Venetian, actually Blazo, because the one in the Venetian closes at 11. Oh my, you know how to live. You know how to live, and I've always said that about you. It's a pleasure to welcome, as our first guest of the day, Caitlin Gordon, who is the director of storage marketing at Dell EMC, and good afternoon. Caitlin, thanks for joining us. Thank you so much for having me. A cube vet, right? You're a cube veteran. I mean, as three, is that like, is you're over the hump of the veteran at this point? Yes, you deserve a varsity letter now. Do I get a letter jacket too? Well, we'll work on that later. I'll give you a cube sticker for now. How about that? Okay, I'll take a sticker. All right, so you've given, you've launched, I would say given birth, but you've launched a brand new product today, PowerMax. Tell us all about that. First off, Payne's the big picture, and we'll drill down a little bit and find out what's so new about this. Yeah, absolutely. So, hot off the presses. It announced just two hours ago in the keynote this morning. So, PowerMax is really the future of storage. You know, the way we're talking about it, it is fast, it is smart, and it's efficient. So, we could kind of go through each one of those, but the headline here, this is modern Tier Zero storage. It's designed for traditional applications of today, but also next-gen applications like real-time analytics. You know, we have some metrics that show us that up to 70% of companies are going to have these mission-critical, real-time analytic workloads, and they're going to need a platform to support those, and why shouldn't it be the same platform that they already have for those traditional workloads? So, let's just go back a minute. What makes it smarter? What makes it more efficient? You know, what makes it faster? Can we start with fast? Yeah, sure. Okay, that's my favorite one. Sure. So, fast. I got some good hero numbers for you. All right, good. So, we'll start there. 10 million IOPS. That makes it the world's fastest storage array. Full stop. No caveats to that. 150 gigabytes a second throughput. We've got under 300 microseconds latency. That's up to 50% faster than when we already had with VMAX All Flash. So, that's great. We get fast as Bob said, right? But, how do we actually do that is a little bit more interesting. So, the architecture behind that, it is a multi-controller, scale out architecture. Okay, that's good. That's a check. You had to start with that. But the next thing we did is we built that with end-to-end NVMe. So, end-to-end NVMe means it's NVMe-based drives, flash drives now, SCM drives, next generation media, coming soon. It's also NVMe over fabric ready. So, we're going to have a non-disruptive upgrade in the very near future to add support for NVMe over fabric. So, that means you can get all the way from server across the network to your storage array with NVMe. It's really NVMe done right. So, let's talk about today. NVMe fabric ready, which I love NVMe over fabric. Connectivity, getting 10 million IOPS to the server in order to take care of that. What are the practical use cases for that much performance? What type of workloads are we seeing? Where we see this going in is to data centers where they want to consolidate all of their workloads, all of their practices, all of their processes on a single platform. 10 million IOPS means you will never have to think about if that array can support that workload. You will be able to support everything. And again, traditional apps, but also these emerging apps, but also mainframe, IBMI, file, all in the same system. So, can we talk about that as opposed to, let's even compare it to another Dale family technology. We just had the team, Shana May, and his VMware customer talking about SAP HANA on ExtremeIO. ExtremeIO is really great for one-to-one application mapping such as SAP HANA. So, are you telling me that PowerMax, this position that I can run SAP HANA in an additional to my other data center workloads and get similar performance? Absolutely, it is the massive consolidator. It's kind of an app hoarder. You can put anything on it that you've got. And it's block, it's file, and then it's also got support for mainframe and IBMI, which there's still a significant amount of that out there. So, that's an interesting thing here, having all of these traditional data services, usually when we see tier zero type of arrays, Dell, EMC had one just last year, there's no services because it's either go really fast or moderately fast in data services. How do you guys do that? Yeah, well, the benefit of where we're coming from is that we built this on the platform of the flagship storage array that's been leading the industry for decades. So, what we did is we took the foundation of what we had with VMAX, and we built from that this end-to-end MVME PowerMax. So, you get all of that best-in-class hardware, that optimized software, but it comes with all the data services. So, you get six signs of availability, best-in-class data protection, and resiliency, everything that you'd need so you never have to worry. So, this is truly built for your mission-critical applications. Yeah, so really interesting speeds and fees. Let's talk about managing this box. VMAX has come a long way from the symmetric days, so much easier to manage. However, we're worried today about data tiering, moving workloads from one area to another. These analytics workloads move fast. How does PowerMax help with day two operations? So, you've heard the mention of autonomous infrastructure, right? Really, PowerMax is autonomous storage. So, what it has is it has a built-in, real-time machine learning engine, and that's designed to use pattern recognition. It actually looks at the IOs, and it can determine in a sub millisecond time what data is hot, what data should be living where, which data should be compressed. It can optimize the data placement, it can optimize the data reduction, and what we see this as a critical enabler to actually leveraging next-generation media in the most effective way. We see some folks out there talking about SEM and using it more as a cache. We're going to have SEM in the array side-by-side with Flash. Now, we know that the price point on that when it comes out the door is going to be more than Flash. So, how do you cost-effectively to use that? You have a machine learning engine that can analyze that data set and automatically place the data on that when it gets hot or before it even gets hot, and then it move it off it when it needs to. So, you can put in just as much as you need and no more than that. So, let's talk about scale. I'm a typical storage admin. I have my spreadsheet. I know what lungs I map to, what data and to what application, and I've statically managed this for the past 15 years, and it's served me well. How much better is PowerMax than my storage admin? I mean, I can move two or three data sets a day from Cache to Flash. Really, what this enables from a storage administrator perspective, you can focus on much more strategic initiatives. You don't have to do the day-to-day management. You don't have to worry about what data sending where. You don't have to worry about how much of the different media types you've put into that array. You just deploy it and it manages itself. You can focus on more tasks. The other part I wanted to mention is the fact that you heard Jeff mention this morning that we have CloudIQ in the portfolio. CloudIQ, we're going to be bringing across the entire storage portfolio, including to PowerMax. So that will also really enable this cloud-based monitoring predictive analytics to really take that to the next level as well, simplify that even more. You know, I'd like to step back to the journey more or less. When you start out on a project like this and you're basically, you're reinventing right in a way, do you set, how do you set the SPACs? I mean, you just ran off a really impressive array of capability. Was that the initial goal line or how's that process? How do you manage that? How do you set those kinds of goals and how do you get your teams to realize that kind of potential? And some people might look at you a little cross-eyed and say, are you kidding? How are we going to get there? I don't know. We always shoot for the moon. So we always, this type of product takes well over a year to get into market. So you saw a PowerMax Bob on stage there talking about it. So his team is the one that really brings us to market. They developed those requirements two years ago and they were really looking to make sure that at this time, as soon as the technology curve is ready on NVMe, we were there, right? So this just shipping with enterprise class, dual port NVMe drives. Those were not ready until right now, right? This box starts shipping next week. They were ready next week, right? So we're at the cutting edge of that and that takes an extraordinary world class engineering team, product management team that understands our customer's requirements that we have today, because we have thousands of customers, but more importantly is looking to what's also coming in the future. And then at some point in the process, things do fall off, right? So we have even more coming in future releases as well. So let's talk connectivity into the box. How do I connect to this? Is this ice guzzy? Is this fiber channel? What connectivity? Yeah, so this is definitely fiber channel. So and our NVMe over fabric will be supported over fiber channel with this array. What we find with the install base with our VMAX install base especially, they're very heavily invested in fiber channel today. So right now that's where we're still focused because that's going to enable the most people to leverage it as quickly as possible. We're obviously looking at when it makes sense to have an IP based protocol supported as well. So this storage is expensive on the back end. Talk to me about data efficiency. Are we coming out with, because a lot of these tier zero solutions don't have D-Doot out the box. Or they have it, but if you use it, you can't actually get the performance that you pay for, right? There's no point in turning it on. Yeah, it's like, yeah, we checked the box, but there's really no point. Yeah, so VMAX had compression. VMAX all flashes had compression. And what we've done with PowerMax is we now have inline deduplication and compression. The secret to that is that it's a hardware assisted. So it's designed to, that card actually will take in, it'll compress the data, and it also passes out the hashes you need for dedu. So that it's inline, it will not have a performance impact on the system. It can also be turned on and off by application, and it can give you up to five to one data reduction. And you can leverage it with all your data services. Some competitive arrays, if you want to use encryption, sorry, you can't actually use dedu. The way we've implemented it, you can actually do both the data reduction and the data services you need, especially encryption. So before we say goodbye, I'm just curious when you see something like this get launched, right? Huge project, year long as you can say, and even further back in the making, just from a personal standpoint, I mean, you get pumped? I mean, I would imagine this is the end of a really long road for you. We have been working, for the marketing team, we've been working on this for months. It is the best product I've ever launched. It's the best team I've ever worked with in the past two days, since I landed here to getting that keynote out the door, has been so much adrenaline built up that we're just so excited to get this out there and share it with customers. And what's this done to the bar in your mind? Because, I mean, you were here, now you're here, but I mean, tell me about this. What have you jumped over in your mind? We have set a very high bar. I'm not really sure what we're going to do at this point, right? From a product standpoint, it is in a class by itself. There is just nothing else like it. And from an overall, what the team has delivered from engineering all the way from my team, what we've brought together, what we've gotten from the executive, we've never done anything like it before. So we've set a high bar for ourselves, but we've jumped over some high bars before. So we've got some other plans in the future. Well, John, I'm sorry to hurt you. Let's not end the conversation too quickly. All right, all right, sure. It's got some burning questions. Yeah, I have some burning. This is a big product. So I still have a lot of questions from a customer perspective. Let's talk data protection. You can't have mission critical, all this consolidation without data protection. What are the data protection features of the PowerMax? I'm so glad you asked. It's been a decade in data protection. It is a passionate topic of mine, right? So you look at data protection, I kind of think of it as layered within the array. So we have very efficient snapshot technology. You can take as many snaps as you need, very, very efficient to take those. They don't take any extra space on them when you make those copies. They cannot use those as tertiary copy stack to point to workloads such as refreshing QA, DAV, et cetera. Yeah, absolutely. You can mount those snapshots and leverage those for any type of use case. So it's not just for data protection, it's absolutely for active use as well. So it's kind of on the array. Then the next level out is, okay, how do I make a copy of that off the array? So the first one would be, well, do that to another PowerMax. So as you probably know, the VMAX really pioneered the entire primary storage replication concept. So we have certainly Async, if you have a longer distance, but a synchronous replication, but also Metro, if you have that truly active, active use case. So truly the gold standard in replication technologies and our customers, it's one of the number one reasons why they say there is no other platform on the planet that they would ever use. And then you go to the next level of really talking about backup. We have built into PowerMax the capabilities to do a direct backup from PowerMax to a data domain. And that gets you that second protection copy, also on a protection storage. So you have those multiple layers of protection, all the copies across all of the different places to ensure that you have that operational recovery, disaster recovery in that array, and that the data's accessible at all times, no matter what the scenario. So let's talk about resiliency. When we go into our data center, you see a VMAX array. There's a big box with cabinets of shelves and you're thinking, wow, this thing is rock solid. Look at the PowerMax. I think it's a, what, about a 6U? I think it's pretty cute, right? Yeah, it's pretty cute. Oh, that's a pretty array. Yeah. You have one over there. So when you see a VMAX, it just gives you this feeling of comfort. PowerMax, let's talk about resiliency. Do we still have that same VMAX rock solid? You know what, you go into a data center and you see two VMAX and you're thinking, this company's never going to go down. Right. What about PowerMax? Guess what? It is the same system. It's just a lot more compact. You know, we have people consolidating from either VMAXs or competitive arrays where they're in four racks and they come down into maybe half a rack. But you have all the same operating system, all the same data services. You have non-disruptive upgrades. You have to do a code upgrade across the whole array at the same time. You don't have to do rolling reboots of all the controllers. You can just upgrade that all at the same time. We have component level fault isolation. So if a component fails, the whole controller doesn't go down. All you lose is that one little component on there until you're able to swap that out. So you have all of the resiliency that over six nines availability built into this array just like you did with the ones that used to be taking up a bit more floor tile space. You know, the PowerMax is about 40% lower power consumption than you have with VMAX All Flash because it can be supported in such a small footprint. So are we going to see PowerMax in converged system configurations? Yeah, absolutely. So if you're familiar with the VX Block 1000, which will be launched back in February, it will be available in a VX Block 1000. Of course, the big news on that is you have the flexibility to really choose any array. So it could be an X2 and a PowerMax in a VX Block 1000. So let's curious, what is the, now that we have PowerMax, where's the position of the VMAX 250? So the, I'm glad you asked. It's an important thing to remember. VMAX All Flash is absolutely still around and we expect people to buy it for a good amount of time. The main reason being that the applications, the workloads, the customers, the data centers that are buying these arrays, they have a very strict qualification policy. They take six, nine months, sometimes a year, to really qualify even a new operating system, let alone a new platform. So we absolutely will be selling a lot of VMAX All Flash for the foreseeable future. Well, Caitlin, it's been a long time in the making, right? Absolutely. Huge day for you. Yes. So congratulations on that. Thank you. Great to have you here on theCUBE and best of luck, I'm sure. Well, you don't need it. I mean, you've, like I said, superior product, great start and I wish you all the best down the road. Thank you. I hope to see you guys again soon. Caitlin Gordon, you know that before. Yes, before. We'd love to have you back. Caitlin Gordon joining us from Dell EMC. PowerMax, the big launch, coming just a couple hours ago here at Dell Technologies World 2018. Back with more live coverage here on theCUBE after this short time out.