 Live from Las Vegas, it's theCUBE, covering Dell EMC World 2017. Brought to you by Dell EMC. Well, welcome back here to Las Vegas. We're in the Venetian. theCUBE is at Dell EMC World 2017. Good afternoon or good evening if you're watching out on the east coast or perhaps overseas. Good to have you with us as we continue our coverage here on theCUBE. I'm John Walls along with Keith Townsend who is the principal at CTO Advisors. And we're joined by Boaz Palgi who is the vice president and GM and one of the founders of Scale I.O. And Boaz, thank you for being with us here on theCUBE. Good to see you. Yeah, my pleasure. First time I think, is that correct? No, no, no, it's my fourth time. Oh, first time with me. Yes, first time with you. Yes, yes, yes. My apologies. We're just about talking about the history a little bit. And I think it really says something about the growth, the explosion of what you've seen. 14 employees back in 2013. Dell makes the purchase. Today you have 220 plus working. So obviously you've had a lot of great growth, a lot of expansion, but a lot of success. Yes, yes. We've experienced a lot of growth, not just in the number of people, but also in the customers. And not just in number of customers, but also in the capacity in production with customers. So today we see, we have well over 300 large enterprise customers like financial institutions, telcos, all kinds of other enterprises. And also on top of that, some mid-sized and even small customers. And what we see is that the capacity sizes of our customers have been growing over those four years as well. So if four years ago we had maybe a part of the storage estate in some of our customers, today we have quite a few large enterprises that have completely standardized their entire block storage estate on ScaleIO. And maybe one example of that was today in the keynote opening of this event, Dan Maslowski of Citigroup presented how they've been using ScaleIO, they're running ScaleIO on tens of petabytes today and still growing very, very fast and with a lot more capacity to be added over the next few months and years. So what's going on though? I mean, why are customers in Citigroup being one who's already made the move but are making the move over to SDS? I mean, what's generating that kind of activity and what kind of gains do you think they're realizing from that? So I think there are two ways to look at it. So one way to look at it is that actually storage arrays were invented 25 or 30 years ago in order to work around a problem of lack of resources for CPU, for processing, for memory in the application servers. So 25, 30 years ago an application server had barely enough resources to run a single application and so if people wanted to add another application to manage the storage, the disk, et cetera, they had to take another server, fill it with disks and that became the storage server which is the storage array of today. But nowadays, application servers obviously have ample resources in terms of CPU, memory, network, bandwidth, you can put any media whether it's flash or magnetic in your servers and so you still have enough resources and that's exactly where Scale.io comes into play so we take those little part of those resources to actually provide enterprise class storage capabilities in a software form factor alongside the applications or the hypervisors or the databases on the same servers. So this is maybe the technology enabling the shift in the paradigm that has been happening. On the other hand, when you look at what is possible today, products like Scale.io make operations of the data center significantly easier. So if in the past you needed to have dedicated storage products that were actually islands by themselves, you couldn't really interoperate between various storage products of various vendors, you needed dedicated storage teams that were specialized on that storage. Every three years, storage estates would come up for refresh, all the competitors would start bidding, you would start getting very expensive and intrusive data migration projects from the old storage to the new storage. All of that is something of the past when you work with software defined storage like Scale.io. So Moaz, let's talk about that for a little bit. Wikibon did research and determined that they call this market originally server sand, some people may call it hyperconverged infrastructure would overtake traditional storage arrays and sales in the next couple of years. Yes. How, you talked about ease of use. Let's talk about deployment. How is Scale.io consumed? So we see several form factors of Scale.io today. So the most obvious one is software. So some of our customers buy Scale.io software, buy the servers of their choice which might be Dell servers or HP or IBM or any other server vendor out there and they build their own state just like they used to buy servers to run their databases for Oracle or to run the operating system, the hypervases, et cetera. Now they also run the storage as another application really on their servers. So that's one form factor software. Actually people, some of our customers today downloaded software from Scale.io from the internet, started to use it, started to grow it and then came to Dell EMC to buy the license and to grow it and to put it in production for real. So that's a weird, that's a kind of strange statement you said. This is a platform that holds petabytes of storage. You're telling me customers just downloaded it and started that, I missed the whole sales process and before the download part. That's very unusual for EMC. It's unusual for EMC and especially for all vendors really in the storage space, but this is a new world. So Scale.io software is freely downloadable for testing purposes and customers find it, download it and we have not a small number of customers that actually came to us that way. Hey, we already used Scale.io, we tested it, we took some servers that we had lying around, we built a cluster, it works, it gives tremendous performance, it's easy to operate, we want to roll it out in production and we understand we need to buy the license for that. So this is one form factor, the second form factor that we see are appliances. So Scale.io obviously supports the 14G servers of Dell. We are agnostic really to the underlying hardware, but this is one of the Dell server approaches that we are supporting is to provide appliances based on 14G servers running Scale.io together with hypervisors like ESX or Hyper-V or KVM and a management software around it that we call AMS that allows customers to manage that entire stack of the server and the scale software and the hypervisor software and the firmware, et cetera, with single clicks configuration, single click upgrade and pretty soon also single click deployment of virtual machines and storage together. So this is the second form factor appliances with a whole management package for the entire stack really wrapped around that. The third form factor that we see are the VX REC flex approaches where VCE or CSPD nowadays are selling entire RECs including networking, compute and Scale.io storage and customers can buy these RECs, plug them in and start running their applications and their environment out of the box. So it's all about simplicity, right? I mean, one click, you're talking about the combination of force right into the new structure. So it's all about making it a lot easier at the end of the day. It's a big problem. Huge problem. I would say it's simplicity of management but also simplicity of operations. So in the past, traditional storage states force people to deal with storage islands, forklift upgrades from all the systems to newer systems. When you have an array that's full, you now need to somehow migrate data to another array. There are a lot of operational challenges with the traditional approaches that completely disappear just like that when you deploy a software like Scale.io which completely scales across all these clusters, across all these environments, across bare metal operating systems as well as across multiple types of hypervisors and you really get one big pool, if you may, of storage that, well, being a big pool also provides among the best performance in the industry as well. And this is because of our architecture that is completely parallelized and it makes it possible to not only aggregate capacity but also aggregate performance across a large number of devices and nodes. So, curious geek question. When EMC originally bought Scale.io, Chastak did a, I think what he called a face melting demo of using Scale.io in AWS and he, crazy, I don't know, it was like a million IOPS or something coming out of AWS, shows the portability of the application. Future of Scale.io, do you see a use case for Scale.io in the cloud? Well, Scale.io, in many cases, enables the cloud. So, we see one of our main use cases is infrastructure as a servers and this is really private clouds in the enterprise or managed hosting or public clouds in telcos or managed service providers environments. And this actually represents a very significant part of our deployments. Another part of our deployments are the traditional enterprise applications like the Oracle and SQL and hypervisors of the world. And then we also see deployments of the newer type of applications like MongoDB, Cassandra, all kinds of OpenStack implementations, et cetera, also on Scale.io. So, I hate to jump ahead, but it's always interesting to talk with people such as yourself, we're always kind of thinking ahead. Like, what's the next big headache or what's the next big problem that you'd like to tackle, that you'd like to challenge, that you think that with a more polished or more defined storage capability would solve whatever that dilemma might be that's emerging for the enterprise? Well, I think the first hurdle we need to pass is just the challenge for most industry veterans in particular to make this shift from the build like a tank, traditional storage arrays that you can touch and see to software that has a connotation or a perception of it's just software, I can touch it, I can see it. How can it be robust? How can it be performant? How can I operate it in an easy manner? As a matter of fact, all of those topics are better with the Scale.io software than with the traditional enterprise arrays. And we know we've built some of them in the past, but in Scale.io, you get the most advanced benefits in terms of operational ease, elasticity, scalability, performance, flexibility of deployment, readiness for the future. So, agnosticity to the underlying hardware, underlying media, and so this really makes a data center a lot easier to be operated and also a lot lower cost because you eliminate a lot of the complexity, you eliminate a lot of the smaller vendors that only deliver a small part of your hardware estate because now you, as a customer, can really leverage everything on your x86 hardware. And this is commodity par excellence. You can go out there, you can get server vendors to bid for the hardware estate that you want to run and on that estate, you can run your applications, your databases, your Scale.io software, whatever you need. So, if you're telling them you can touch it, or you don't have to touch it, you don't have to feel it, trust me, it's real, right? It's software business. Well, don't trust me, try it. Try it, okay. Thanks for being with us. We appreciate the time here on theCUBE and good seeing you. Again, fourth time around, we're about to join the five-time alumni clubs. Congratulations on that. Soon, yes, yes, yes. Caleb, by the way, tell you a little bit about the version, the new version. Oh, please, yeah, we are new back, I'm sure. So, maybe very quickly, end of this year, we are releasing a new version of Scale.io, Scale.io.next. The main items there are we are delivering space efficiency, but we provide it in a dynamic manner. So, one of the big downsides of putting space efficiency in storage systems is that it usually hits performance quite significantly and especially if the data is not too compressible. In Scale.io, we will dynamically compress data based on the compressibility of the data. So, if data is not compressible, we won't waste resources trying to compress it. If data is very compressible, we will use more resources, but we will also compress it and we will be able to do that with a very little, very small degradation of performance compared to the non-compressed environment. That's one, two, we introduce volume migration in a non-disruptive manner, enabling customers to move volumes from flash-only to magnetic-only to hybrid environments on the fly without any disruption to the ongoing applications. We introduced a support for VVOLS in order to be able to run all the Scale.io capabilities, all the Scale.io volume capabilities on a virtual machine granularity level in ESX. And we are also introducing the next level of the AMS management software, which is the wrapper around the server with Scale.io software appliance bundles that enables you to manage the entire stack. Timing again? End of this year, end of 2017. Good deal, you got a full plate, don't you? We do, we do. Very good, well done, thank you again. And sorry about that, I know you had news you wanted to get in, I'm glad you did. Well, thank you for the opportunity. You bet, all right, back with more of theCUBE here from Dell EMC World 2017, right after this.