 from Las Vegas. It's theCUBE, covering AWS re-invent 2018. Brought to you by Amazon Web Services, Intel, and their ecosystem partners. Welcome back, here on theCUBE. We continue our coverage at AWS re-invent. We are at the Sands Expo. 40,000 strongs, maybe more attending this year's show. And once again, this show floor is packed and it is a quite impressive display. We've got a lot of great exhibits, a lot of excitement in the air here. Along with Justin Warren, I'm John Walls. Again, welcome back here on theCUBE. We're joined now by Prakash Darjeh, who's the general manager of FlashArray at Pure Storage. Prakash, good morning to you. Yeah, good morning. Glad to be here. Thank you for being with us here. All right, so you've got some, an exciting new direction that you guys are going, looking at Cloud Data Services. Tell us a little bit about that, about your on-train to that world. Well, it's interesting. Most people know Pure as an all-Flash company. That's how we started. But if you actually take a look at what Pure's really great at, is it's building software for platforms or technology that might not be mature. So Pure made an early investment in consumer-grade Flash when the market was going to Enterprise and said, you know what, consumer-grade Flash doesn't have the resiliency or the enterprise-grade characteristics that people need. And in that early entrance, we built software to deal with the crappy hardware, basically, at the end of the day. And that's generally worked out well for Pure in the all-Flash market. And what we realized is the same value propositions we were able to build around higher performance and reliability and enterprise-grade characteristics were some of the characteristics that were missing on another platform that we saw, which was cloud infrastructure as a service. And circling the show floor, it's interesting. I was talking to some of the customers and I'm hearing a lot of feedback around the, what are you really doing with our cloud data services? And really what we're doing is we're trying to say, you know what, you shouldn't have to compromise between on-premises and cloud. You should get the same enterprise-grade characteristics you have on-premises in the cloud. And frankly, if the API and the software is the same, then you can lift and shift and move back and forth. So from a value proposition standpoint, cloud provides you the instant available capacity and agility. On-premises typically has been optimized for a high degree of performance, cost, and resiliency. And now you have the ability to add agility to that angle and start anywhere and move anywhere. That's really the goal of what we're trying to do with our cloud data services. That is a theme that we've been hearing so far in the show over the last couple of days, that there is this realization that workloads can live in multiple locations and that maybe the cloud isn't right for all of them or maybe it's not right right now or maybe we try something in the cloud and then we actually want to move it to somewhere else. So being able to do that is what a lot of enterprise customers certainly want to do. And we're hearing from a lot of vendors that that's what they're trying to enable. And it sounds like that's what you're trying to do here with Pure is opening up this new avenue for well, you like Pure here on site. We would love to use some Pure over there in the cloud and now you can. Well, that's one part of it because people always have to like, when you're making a decision, you have to decide where you're going to develop. Am I going to develop on premises or am I going to develop in the cloud? And typically I like to liken it to center of gravity. Where's your center of gravity? And data has a lot of gravity. So if your data is primarily here, that might be like, hey, I'll do it, develop here. If it's something new and you don't have a lot of data gravity, you might decide to develop in cloud. But increasingly we see applications being hybrid applications. For example, today, Salesforce CRM is a SaaS application. You can argue that that's completely cloud, right? But anything you sell needs to book in a finance system, most of which is on premise today. So the application workflow crosses both anyway. The data workflow crosses both anyway. But in IT management, it's just different across both of those worlds. So we increasingly see the need for hybrid applications where you can use the best of what's available where. If you want to use AI algorithms in one cloud and you want to use office services from another cloud and you want to use infrastructure build services from a cloud and data from a non-premises system to go ahead and orchestrate and build your app, why shouldn't you be able to? The only way to do that is to bring the application architectures together and between VMware cloud and Kubernetes that's starting to happen. But at storage, no one's really bridging that divide in terms of making storage look the same on both sides. And that's what we're doing. So the big challenge with storage is that everyone knows that state management is hard as well, but being able to move that data, like you said, it has gravity. What if I make a choice today and then the pace of innovations is so fast then I'm going likely to need to change my mind later on and I'm going to have to move data around. How do I do that? Is it, how do I, if I've chosen to use some pure in the cloud and I want to bring it back on site, how would I do that? Well, so it's interesting. You went like there's multiple ways to move the data. The challenge isn't actually in the data movement itself. It's whenever, because the data has gravity you always build things around it. Meaning you have applications sitting on it. You have interfaces connecting to it. You have workflows such as I'm doing development and I have APIs that are spinning up new volumes. All of those workflows and all of those integrations have to be redone if you want to move it. Like moving the data could be as simple as like dump it to a file, ship it over there and upload it. You know what I mean? And there's more sophisticated ways to move. So the data movement isn't the challenge. All the integrations and workflows you build around your data is. So really what's most important is ensuring you build a consistent API across both environments. So the way we enable that today is we've taken the same pure software that we've built and optimized for a flash or AM or an X and we've now optimized it for a third platform called AWS Infrastructure. And we'll probably do a fourth and a fifth if you read the tea leaves into the future as well in cloud environments. But that software is the same. Meaning like I met a customer that was interested in, they've built on AWS today. But they have online curriculums for college education and they have to take snapshots for curriculum development every semester that they send to multiple locations to build coursework. And what they're planning on doing now is setting up a direct connected in Equinox flash array that they're basically synchronously replicating between our cloud block store and AWS and taking their snapshots from this environment because they're space saving snapshots and they get to save on the export taxes. So when you treat the software the same, it's amazing how people will start using it because at the end of the day, your orchestrations, your APIs, all of your workflows are the same. So now you want to move, there isn't a tax to rewrite anything. You just move the data. And once we add other platforms, then you have the ability to use the best capability that's available where. Is there any kind of a danger or, I wouldn't say danger, may not be the right word. The fact that you can make these transfers relatively frictionless or at least a lot simpler, a lot more convenient now. All of a sudden, I want to move everything. I don't know. And I'm not as selective as I might have been before. And I'm just take it and dump it and move it. And I don't have to identify what's really necessary, what's valuable. Instead, I'm just taking the simple way out as a customer, as a client. When you coach them along in that way at all to help them prioritize. So just because we can do it, doesn't mean we have to do it. Yeah, we've started getting into that discussion around education. What's interesting is, as we've been having these customer discussions, we find that the level of education in the cloud is pretty disparate. Some people who are very like cloud first, I'm going all in, now know the challenges. And we don't really have a lot of education to do. They've got a cost model. They've got performance comparisons. They've got reliability comparisons. So they know that from a performance cost reliability standpoint, having control over your own infrastructure provides the most control over those elements. But they know that's not the most agile way to do things. So they're treating the public cloud as a instantly available agility capacity. And as they mature, they're moving certain things back into more hosted or private. On the flip side, we have other customers that have started on premises, even a lot of our own customers who are using FlashArray in a hosted way. And they're saying, you know what? There are certain workloads that need to sit closer to different locations in the field. We don't have the networks for that. So we're going to actually leverage the public cloud for that. And given that we can move it, we're going to do that. So what we're finding is people that are educated are making these as very conscious decisions. I find that the market that is uneducated is an interesting market where, you know, I've met a customer a few weeks ago in the oil and gas, big oil and gas company that everything's going to the public cloud. But they have nothing there right now, right? And this, you know, like any hype cycle, it's like, hey, we're going to do this ever. We're going all in. And I'm like, have you thought about this, this, this, this? And they're like, no, we're going all in. We want to get out of data centers. So, you know, we're like, okay, we'll support you in that journey, but we're going to guide you in terms of like, hey, it's a deep pool. You probably want a floaty or two, so you don't sink. And we're giving you the floaty. It's a good analogy. I love the analogy. That's great. Prakash, thanks for being with us. And I assume you brought some floaties with you in case you need to hand them out on the floor this week just in case. Yeah, we've got peer storage here. We've got a nice presence where we're handing out some nice swag like everyone here. Excellent. Good deal. Well, thank you for being with us. We appreciate it. All right, thanks for having me. Prakash, joining us here for peer storage. Back with more at AWS Free Invent. We are live in Las Vegas and you're watching The Cube.