>> Announcer: Live from Las Vegas, it's theCUBE. Covering Dell EMC World 2017. Brought to you by Dell EMC. >> Welcome back to Dell EMC World 2017. We're live here in the Venetian in Las Vegas. Day one of the three day show. Had Michael Dell out on the keynote stage earlier today. Also had David Blaine, world famous magician. Pretty interesting performance to say the least. >> Yeah I went down to get an ice pick. (man laughing) During our break. >> We'll get into that later but it was interesting. Keith Townsend, John Walls also joined by Itzik Reich who is the CTO of XtremeIO at Dell EMC. Itzik, thanks for being with us. It's good to see you sir. >> Thank you very much. >> All the way from Tel Aviv and great to have you. Alright, so your sweet spot of the company is giving birth to a new baby today. >> There you go. >> XtremeIO X2, tell us about that. What spawned that, and then what that responses be, what you developed. >> Right, I think in order to understand Xtreme, you need to start with the beginning, the X1. So, November 2015 I was having my class reunion, meeting my ex-girlfriend, and we've launched X1. And X1 became, within two quarters, the largest sole Dell flash array in the world. From nowhere to the largest sole flash array, at least in terms of units sold to the market. Right, both Garthner and I. And it was huge. A huge building and a success for us. A success because nobody would become the number one leader. And we built them because we didn't have the life cycle to normally mature a product. Right, so you mentioned being a father. I'm a father to two daughters, lovely daughters. One of them is six years old, one of them is five. And the young one is starting to show some signs of being a really clever person. And I'm afraid that somebody will tell me, oh she can skip the first class. Because skipping class serves some association with it. Social aspects of it. So we've been really busy trying to understand XtremeIO X1. Making super stable. Today we're already about 5/9 in the market. But it also would stand to refresh the product and come with something new. So our life cycle wasn't a traditional year or year and a half of refreshing the product. It took us longer for us to X2 and this is what we announced today. So what's new with X2. The first thing is the ability to come with really Dell's XO Drive and Dell's configuration. In X1 each DAE, you could put up to 25 drives inside of the DAE. And X2 can put out up to 72 drives per DAE right. And you can scale just like before. Up to 8X bricks. It's a huge capacity which you need for the vast majority of the use cases out there they don't know. Just VDR or just a single database is right. Today XtremeIO can fill pretty much every transaction while closing including virtualization wall close. You just need a lot of capacity for thousands of VM's. So that's one of the things. The other thing we improved performance of the X2 array. And the magic story around there was that because of the thousands and thousands of customers that we're involved with really got the good insight of the workload that they are running. And what we found out is something very interesting. The majority of those customers are running workload that they're very small block size. So you storage every item that arrives in the system as a different blocks characteristic and we found that the majority of them are using very, very small block size. And we want them to improve the performance of those block sizes. The IOPS and the latency. And we also wanted to make sure that it's actually more economical cheaper than the very expensive drives that the new NVMe drives that are out there. So different design goals. Making it faster and also making it cheaper in different dimensions. So we come with a new feature called Drive Boost. In a nut shell, in a nut shell Drive Boost will give you 80% better latency for pretty much every walkthrough that is out there. >> So... With that small block sizes versus big block sizes. Why is that important? We're at a conference and we're talking a lot about digital transformation. CEO, we teased John earlier. You know he's a sports guy, he doesn't do LAG goals. >> (laughing) Sorry. >> That's alright. >> Help us understand the value of that data type. >> Sure, so you know we like to think about digital transformation but at the end of the day. You're the customer, you have a database. You'll use it on query or queries against the database. If it's a very large database, there are thousands maybe even millions of queries everyday. Those queries take time for the end user to get a response for. So let's assume that you want a monthly report. And this report normally takes nine hours to generate. If I can shrink the report crunching time to two hours instead of nine, that means that I have provided better value for the business success. Right. One of the stories is that we have a financial customer in the Middle East. They need to generate the report every month between midnight because this is where they locate their reports. Up until eight o'clock in the morning. Why eight o'clock because this is when the employees start to come to work. And every hour that they exceed after the eight hour generation they get fined by the government. So if I'm saving this customer four hours then they are not getting fined by the government for generating the report. That's a true value for the customer return. Cause those things are important. People tend to think about just performance numbers in terms of IOPS but the real magic number is latency. How quick can you make the query? Whether it's a database application or a VDI VM or just a generic web server running on a Voltron machine. Those are the important things today. >> So transactional apps. Big deal. Are these transactional apps, we learned a lot about virtualization and cloud computing to date. Are these transactional apps running in a virtualized environment or are we still relying on big heavy metal workloads going to treat IO2. >> Yeah, it's a good question. At least from my experience some would argue that anywhere between 70 to 80% of the customer that allowed it went full virtualized. So their running their entire application running either under V6 or a Microsoft type of V. So they are fully virtualized. Some of the customers are still running their workload on a traditional physical servers right. Even in the S6 at the end of the day it runs on a physical server to all day the kill in itself. But yeah, the majority of them are already there in terms of virtualization. >> So what are customers really excited about when it comes to features sets for an XIO2 versus XtremeIO version wise. >> Right, amazing question. So performance, we've already discussed performance. 80% better latency, that's not something that you get because of the usage of better CPU's. Intel moves slow, it's basically dead right. They don't give you 200% performance between generation so we wanted to do something else and solve the same problem. The other thing is quality of service. We are not cheaping NGA but it's coming soon. The ability to give a specific VM, a specific IO copying and the latency copying. And also could give you the ability to burst to more IOP's techniques needed for a couple of minutes. So quality of service I the noisy neighbor right. Somebody generate too much noise you want him to be quiet. That's what quality of service is. The other things that we've announced native replication. We found out finally of our own replication that can replicate between one XtremeO2 and another. But it's not a traditional replication. The unique thing about XtremeIO was always the cusp. The content of dressable architecture. People typically think about it as a D Duplication feature but in fact we don't have a feature called D Duplication. We analyze the data as it goes through the system itself. And we give a unique shot signature to each one of those blocks. And if the shot signature already exists in the system we dupe the block. But it's not the feature per se. That why the D Duplication's so fast on XtremeIO. So up until now the customers architecture was only applicable to writing the data into the array itself. Now it's also applicable for replicating the data. So for example if you have a data reduction of five to one which is very common in virtualize use case. Many VM, many the same template and so on. You know need to replicate four times less the data at the source to the destination target. Right. So that's a very, very big thing because you need to replicate more and more data. But the 24 hour window isn't changed. God didn't upgrade it where the server respects the time. Right. >> (laughing) Right. >> It's still 24 hours per day. So this is super important for us and we're very excited about it. And the other thing is that, again larger denser configuration of the array itself so the customer can have up until two-thirds cheaper. The drive, the cost drive of the XtremeIO in itself so it's cheaper for them to put their walkthrough on ExtemeIO. Whether to really pick up the just the database that needs all the performance in the world. So we can really become a true enterprise array with those features. >> It seems like it's got to be for you a constant chase though right. You're looking for higher performance, you're looking for lower costs. You've said you just gained 80% increase in your performance capabilities. >> Yup. >> And now people are going to be looking at you over the next Xtreme and so what next? You know, where are the gains to be had in the next generation of technology and just in terms of philosophically approaching that so what do you do. >> Yeah, yeah again another good question. I actually gave a briefing about it just earlier. So, the first thing we need to do is an industry not just the daily insists to lower the costs of the drive itself to be even cheaper than and economical drive. That's not Dell today right, the hybrid mechanical drive. You can get a more economical drive if you apply data reduction on it right. So if you're five times cheaper because of the data that's gets integrated into the array and get a different compress and different provisioning. Then you can be on par with the mechanical drives. So first we want to be on par if not cheaper. We want everybody to move to S's. And we were the first twirl for charade the portfolio of Dell EMC. That's the first thing. The second thing is to really get a better insight into your wall application, wall close. Today people analyze things like IOP's and latency but what does your application really think? Where are the cues in the application stock itself right. How can you find them out in the storage sub-system itself right. So we are on a journey to over there with our importing mechanisms. So a year and a half ago, we started a new project to completely change the reporting mechanism of the WebUI. The interface of XtremeIO right. And today you can really get to drill down into pretty much every aspect. Up until now you had to purchase a third party software that will analyze your walkthrough for you. So things like Instagram, IO's, block size, read and write like I can see pair of blocks. So you can really understand your workload. We also give you something like abnormalities. We can tell you every week this application is being fine but on that Friday for some reason the response time wasn't that good. You should go in and check it out. Maybe it's in the application there is a bottleneck. Maybe it was a bottleneck in the storage load. So you can actually find it out. But I would argue that the long term goal. That's a vision right? That I'm not announcing anything yet. Is really the ability to marriage or combine between the softer defined wall right. The input converge mechanism, to the traditional arrays right. Although SSD's not that traditional. Maybe you can have a denser configuration with very small to DAE but the performance aspect of it will not be drive from the DEA where it actually store the data but from Voltron machines. That you can spin up and down in a cloud like fishing. That will bring you all the performance that you need. That's a thing to me the only gray. The really merging between the walls. Cause there isn't one perfect answer right. The softer refined guys will tell you everything should go to softer defined storage. We will tell you everything should go to flash arrays. But really the truth is like always right in between. And this is really one of the direction that we are approaching. >> I tell you what, for now I want you to enjoy X2 for now. How about that. >> That sound good. >> It's a good day for you. And don't let that five year old skip either. I think that's a good idea too. >> Very good. Very good. Thank you very much. >> And so thanks for joining us. >> Thank you. Thanks. >> Back with more here on theCUBE. We're live in Las Vegas at Dell EMC World 2017. (exciting techno music)

>> Announcer: Live from Las Vegas, it's theCUBE, covering Dell EMC World 2017. Brought to you by Dell EMC. >> Welcome back, here at Las Vegas at The Venetian. As theCUBE continues our coverage of Dell EMC World 2017. Along with our co-host, Paul Gillin, I'm John Walls, good to have you with us. As we, I guess were coming down the home stretch. >> Paul: We are. >> Day one. >> Paul: End of the day. >> Here at Dell EMC World 2017. With us now is Chhandomay Mandal, who is product marketing director at Dell EMC. Double dipping on us, we just had you on a few moments ago. This is nice, we get two shots. >> Thanks for having me. >> Good to have you back with us, yeah, it's good. >> Chhandomay: Thanks for having me. >> So before we were talking about XtremeIO, what you were doing in the healthcare space. Moving over now to copy data management, different part of your portfolio, and kind of what's up in that world with you. So just give us a little rundown, an overview of what you're up to right now. >> Sure, so let's start with what exactly is copy data problem. Why it's a problem, and why we need to be solving it? If you think of any business application, it comes with its production data. But for every bit of production data, you have many different copies. For example, when you're developing applications, you need copies for your development and testing. You need copies for your backup. You need copies for running your analytics environment. It's for every single production database, typically, we see five to 12 copies of that data. And in fact, I did see estimates. The copy data sprawl is like 40 billion dollar market, and 60% of all the data that exists are on copies. Now, our mission, especially with Dell EMC XtremIO, is to solve that copy data problem, giving the customers back a lot in terms of the storage efficiencies, and not only is the storage TCO, but transforming the business workflows. We did copy data management so that they can realize storage and infrastructure settings, but also the business impact from transforming the application workflows and bringing new production market in a much quicker way. >> So you said a 40 billion dollar market, I mean, what are the costs here? Is it just storage cost? Is it bandwidth, is it errors? Lack of data being out of sync? >> So the cost here has multiple components, right? First of all, there is the cost of the lost storage where you need to put the data on. But then, there comes the cost of managing the storage. How do you figure out where you backup copies are, if you need to restore, where are you going to get the data from? It's a cost of inefficiency meaning, like if your developer who is the highly paid, highly productive guy, supposed to be, right? He is waiting for the DBA or the storage admin to give the copy that he needs, then, that's just enough money, right? It's not just the infrastructure cost, but also the soft cost of, like your ability to bring the product to the market in a quicker way, addressing your customer needs in a quicker way. That acts up and those are the components in, like, how I value this intermarket. >> I guess what I'm hearing here, if you got five to 12 copy sets of data, I mean, massive amounts of waste in some cases, right? And maybe some of your clients, they like to know where everything is, but do they lose track of it, and so it's taking up space, taking up money, taking up time. Is that, are these the problems they're facing? >> Chhandomay: Yes, yes. >> Alright, so then, what's the answer then in terms of this better identification? There's X2, get to the heart of that, and help them in terms of better efficiencies? >> How do we achieve that efficiencies? Now, one thing is, the way, first of all, like, if you can consolidate your copies into one single platform. And copies are duplicate bits, right? So, first of all, the first thing in the process is you eliminate all the duplication that exists in your storage. You have your production data base, and you have your copies, which are, if not unique, then basically should not take up any extra space. Now, you take those copies, and make it like a repair plus one. So for example, your Dell part can run tests on it. So when the rights are coming in, only the changes that are happening, that should go into the storage. So that's Part A. Part B is, when you are running production environment, as well as what works on your copies, you need huge performance with consistently low latency. Because you cannot impact your production SLS. You have to meet that. >> You can't tell it, "Hang on, I've got something "else going on over here, right?" >> So you need a platform that can handle consistently high performance with low latency no matter what workload you are running. And then the copies themselves need to be very efficient. They should not take any extra space, unless there is something unique. And they should be able to perform just as well as in a production value. The hard part of this is, you need to orchestrate the inner process, right? I mean, you as in oracle admin. You really do not need to worry about how and where the storage is going to be saved for your copies. You click on a button, and it should do all the steps necessary right from your application console down to the storage. So this is the application orchestration that we in-built with AppSync and XtremIO. And then we have APIs that our customers can use to provide their own service catalog. So using these pillars, we consolidated all the copies, on the same platform, running different applications, with the same SLS, okay? And that kind of helps the customers to bring product faster, and address the copy needs. >> Now, this is a very hot market right now. And I'm thinking there's some startups, I'm thinking of Actifio and Catalogic in particular, that say that you shouldn't have many copies. You should have one copy, and then you should have pointers to that. What's your opinion of that? What are the pros and cons of that approach versus yours? >> So our approach, essentially, I mean, since you mentioned, right, there are copy data management vendors. What they're doing is, you have your production, then you make a gold copy off your production, and from that gold copy you run off different applications on those copies, right? So here you are introducing another element, another software, and another appliance, so to speak, to manage the copies. What we are doing, is kind of like you don't need that extra copy that your analytic part provider can provide. And then there are performance implications with the integrated copy data management that we are referring. The reason we can do it is, all of our metadata is in memory. It does not consume any extra space for storage. And no matter what the workload is, we can offer consistently high latency because everything is, the metadata is operating from the memory itself. So the way the third parties are doing, we do it the same way, even better, and at the production level. >> Another thing, and forgive my technical ignorance here, but David Fleur at Worky-Ban, has talked a lot about the benefits of flash storage. In that you don't have to create copies, you can create a single copy in flash, and then multiple users or applications can work from that. Do I have that straight? He says that's a game changer. >> Yes, that will be that game changer, and that's really like what we do. The caveat to that is, when you are creating the copies, and you want to run applications on the copies, your production should not be impacted, and the copies should also be able to deliver the same performance. And that part has been the challenge with other solutions in terms of providing the same performance, the same data services on the copies themselves. That's the idiot we solve will our intelligent content error in memory metadata architecture with XtemIO. >> You're talking about the integrated data management just a little bit ago. I mean, from a real life perspective, can you give us an idea about maybe a success story, somebody that you can point to and say, "This is how they incorporated that "into their process, I see it work for them, "and we can make it work for you too?" >> So, I'll give you an interesting statistic. We have 3,000 plus customers running XtremIO in production, and we get all the phone home data at our end, and we can see what they're doing. Now, for XtremIO customers, 56% of the copies that they're making, they are running workloads on them. They are not just for local data production. And, all the IOs, XtremIos that is out there in the field we'll see, 40% of the IOs are because of the copies. So we see across the board on the customers. I have many examples. For the sake of time, I'll just speak one. We all know Moen, they are the leading, not American manufacturers of the faucets, right? It's a big shop, and they have like, a lot of SLP landscapes in there. Before XtremIO, they could not keep up with the backups and the copies that they needed. After moving to XtremIO, now they can actually take the copies of their production SLP landscapes twice a day. They are quietly running reports. They are actually running like 90% shorter, and in fact, we were talking with Harvey H., literally, like before this segment, right? He was also talking about how efficient their copies are. I was talking with Scripps Health, who are also going to be presenting in here. They run like 3,000 copies in their environment, with XtremIO and AppSync, and like it's all working great. No impact on the performance, and they are meeting their SLS. >> Well, your performance on theCUBE has been outstanding. Back-to-back saves, we appreciate the time. Chhandomay, thanks for hanging with us. Best of luck down the road, and continued success here at the show as well. >> Thank you, it was a pleasure. >> We will continue with more from theCUBE here in Las Vegas. We are live at Dell EMC World 2017.