 So wow, this is quite an event isn't it? Yes. I've been saying very un-Hitachi like you know this is great marketing here It's you don't really expect that from a touchy, but we we sort of seen a transformation here, haven't we? I think a bunch of companies are having to go through transformation in the storage industry I mean if part of the whole message is that it's not just a sort of standalone device a peripheral But it's part of your data center solution You know you really got to send the message that you're a full partner in the data center And that includes being a lot more visible about customers and use cases and all the other pieces that go with it You're seeing a real transformation in the industry minutes You know we always talk about transformation, but this really feels like it truly is one with whether it's cloud and VMware and Virtualization, you know, what are you seeing Laura? Well, I mean I think the thing that HDS has gotten right Maybe more so than their peers is as the data center is transforming It's also the IT domain that's going through a transformation as well So certainly, you know as data centers move to become more virtualized more scale out architectures as service automation becomes more of a critical role Increasingly the IT department is being asked to manage more than just the infrastructure and the thing that I think HDS has gotten Exactly right is it's now about managing the information and you know They have a lot of capabilities that help them in that front We still have a long way to go as an industry But I think that's a differentiator for HDS relative to this whole data center It is different than what you hear from a lot of other other suppliers We were at Oracle open world last week and the message is we want to own everything yet We have to own the data own the database own the infrastructure and HDS taking a little different approach Laura You said it just manage it. No more facile approach So what's your take on this announcement Laura from a software standpoint your software expert and you see all the different players out there What's what are the highlights from a software perspective here? Yeah? I mean, I think certainly their Their page or record level automated tiering is a differentiation for them is a compelling capability, but their whole unified management strategy I mean in storage software is really going through a transformation from a use We see consolidation of the storage management layer and the device management layer and really what that means is things like Automation or increasingly more important things like unified management across different platforms or across different data types is more important And and so their whole unified approach unified management approach I think is really you know a key component for their software strategy and you know They've added enhancements around things like you know installation and improved workflows and wizards But but really the unified approach I think is right on the mark in particular as it relates to the data center transformation They need to take that Unified management integrate it up into higher levels of that infrastructure staff. Yeah, I was impressed with the unified story You know, you don't expect big Hitachi or LTP to be talking about that unified approach I think that's a differentiator. Don't you Rick? I do think that is is one of the areas where they've needed to work to actually increase their vision So you think they were a little bit behind there needed to catch up I'm not sure they were behind technically the challenge of course is being part of the unified message and being in the game Yeah, and clearly, you know the message is we've seen on the unified side Well server virtualization full props for getting people to start thinking about it is it was much more of a you know Unifying their I think you know Hitachi has now put together It's not just that we're talking to game, but here's our design points You know, we recognize that efficiency just doesn't just mean the software and a thin provision It means you actually start designing systems that are more efficient to deploy and manage and run and operate at the hardware level too So so I think they're listening to the full message and being part of the unified story from that standpoint Are you guys hear a lot of talk about silos and information silos and infrastructure silos and application silos? With the users that you talked to where the silos starting to come down and or is that still a big impediment to progress? Around virtualized infrastructure. What are your thoughts on that? I think it's changing, but I think it's evolutionary. I don't think it's like stair-stop functions I think it's low change over time, but I do think it's starting to change You know one of the customers that we just talked to In the one of the keynotes there had talked about you know His goal is to try and eliminate asylos and not to just do things like tearing within the box But across the box and ultimately tear across the data center. So those are some of the high-level objectives But there's a lot of tearing in the box. Yeah So but you know You know so there's certainly a lot of legacy infrastructure out there and so it's going to be you know an evolutionary process That's my take on it Rick Rick We used to when we when I was at IDC you were a networking expert and now you've come come to the storage side Which is timely right because you're seeing all this talk about convergence, you know What's so what's your take on on the silos and convergence and the whole networking and storage thing? I do think one of them and it has been building over the last five six years And I did join the storage team at IDC partly to deal with how is networking going to change the world of storage? And in a lot of ways what what it's done is is saying Some storage is now in a more independent activity It's it's and it is that transition from storage to information management that what we're really trying to say here and this is almost the new silo is I have a group of people whose their job is to make sure that the information is you know captured is Protected is made available and is you know archived over the long term and that's really their job. It's not to Provision a disk drive or provision alone behind some application So I think that's the sort of new vision What convergence is about is to get them out of that other job out of the sort of point and click and? Rebuild and reset everything storage has it network has it servers have it even the applications have it I think we want to tear the silos down at that level But I think you are seeing I don't want to call it a silo But but two new use cases one is information is clearly that information centric workload There's class obviously the classic transaction centric workload and and definitely Oracle and Oracle world what they were talking about is That's not going away, and that's a specialty area, and and you want to at least think about how you're doing that effectively I think the interesting third one is is is compute is analytics And and that that side is a big data, but big data processing I want to crunch it often and frequently to do something with it and and part of the challenge for the storage industry is Is to basically say we're the ones who are tasked with ultimately protecting the information across those three different silos archive compute transaction and Managing the movement of data between those pools of compute resource or or or storage resource That's going to be the interesting development And I think some of the things to touch is working on here are part of making that their job So you think the storage guys are going to work for the networking guys eventually? You know this is that whole cloud debate in a sense of who works for who I'll be honest and saying I think a lot of the Networking guys are going to be a lot busier figuring out how we You know support all these you know iPads and smartphones and everything else out there and controlling it Keep the hell away from our data. I think the network inside the data centers Evolving into an IO management system So we may not call networking companies the way we do today, but it's the people who run and own the IO So Laura, let's talk a little bit more about this announcement And even a company that's 97 billion dollars in revenue can't do it all so I want to talk a little bit about What's missing big Rick just mentioned, you know big data. How about data reduction? That's something that I I didn't hear you know strongly in this announcement Although they do a lot obviously with things like thin provisioning or what they call HDP Hitachi dynamic provisioning. Do you think that's there's an opportunity there? There's upside there for these guys Yeah, absolutely, and I'm sure they have that in their strategy or in their plans But storage optimization technologies at large are really kind of a critical component to storage architecture and purchasing decisions for customers Because when customers are trying to do more with less right and trying to scale their storage while keeping their their administration in check Their budgets aren't scaling with the same level of capacity growth So optimization has to play a key role deduplication is an area where we see you know, we just saw it, you know Really proliferate in the backup space backup data archive data And now we're starting to see it move more into the the primary data, you know on the file side in particular So there's certainly an area of opportunity, but it's not it's not did you it isn't the silver bullet It's in conjunction or in concert with other optimization technologies than provisioning space-efficient snapshots compression technologies and So yeah, I think that's certainly an area of opportunity or future expansion. So let's talk about that a little bit more It really isn't part of this announcement today, but it's an interesting topic. I know you have a perspective on it I'm interested in your point of view. I'm sure listeners would like that So what does it's how'd you do there? Did they partner with like a Falcon store? I saw it in here. That's they're really showing a VTL, but that's a possibility Do they develop their own did they we had Jack domain on it? didn't I didn't get the sense that they were going to go out and make a bunch of acquisitions necessarily but You know or do they just roll roll their own in-house is it already? Is it are they are they late? Are they too late? I don't think they're too late. They're certainly later than others, but You know if you look at statistics our show our research shows that an average of between, you know 30 and 34 percent of all the total disk storage capacity on the floor is allocated for some kind of backup and recovery function So that's a lot of data that continues, right that continues to grow and backup is highly redundant So it's certainly an area of target for dedu They have a partner they have a partnership with Falcon store around deduplication with the Falcon store VTL They've also had other relationships. I think data protection and disaster recovery is an area of Further consideration for HDS as it you know evaluates where it wants to be in that in that opportunity But clearly there's a lot of opportunity as it relates to selling solutions for those particular workloads whether that be Data protection disaster recovery business continuity archive and dedupe is you know a critical component there But to your point, I think you know they have historically relied on a lot of partnerships and You know probably behooved them to to have a more organic approach, but it's not done yet that whole space We saw a sync sort and that app just do a deal acquisitions There was an article on silicon angle this morning about store-wise supposedly IBM is dumping the store-wise brand and they're going with real real-time data compression a real-time compression Is is what they're saying? It's very very IBM like marketing I guess take the sexy name and call it what it is, right, you know But but you're saying that that whole space perma bit albedo HP. What do they call their product sure stores? Yeah, so so you actually are seeing some more innovations there sure so these guys Hitachi has some time maybe to I think so I think it's still really does in particular as it relates to the primary storage to do for primary stores super early days on that front so There's a lot of opportunity for for innovation there and you know Some companies have multiple dedupe technologies, right? So you look at IBM they bought diligent and then they bought store-wise EMC has avamar and data domain, right? So there's a lot of different ways in which to go about Reducing data footprint right so Rick what um, what do you guys telling customers? You know, what's your big advice when you think about this whole cloud convergence storage changes? What's the advice to customers? Well, there's there's a couple different layers of it I mean obviously at this sort of tactical layer One thing that we clearly do still have to tell customers and surprisingly is virtualizer storage You know, it's not just enough to network it You really do have to put in the facilities to you know do the thin provisioning to to make that those moves ads and changes Basically transparent Not just serve a virtualization think about don't forget the story Absolutely I can't tell you the number of people who we talked to three years ago and even today who you know If they're under if they're less than a year into virtualization, there's a next storage isn't a problem It's there we don't worry about and then after that, you know after they hit that thousandth virtual machine Or you know all of a sudden it's like storage is the bottleneck storage You know either we're rebalancing the systems all the time or the backups are blowing up or you know We just don't know how to scale So you know doing the full virtualization and recognizing that that virtualizing the storage is a key part of it It is a big part of the message The other thing we do say at a higher level and we have to go back is this convergence discussion a little bit is is You know, you really have to rethink a little bit of the it structure You know, there clearly is going to be what we'd call the information facilities management team And today that's the classic IT folks that we think of admins. They have to evolve though You know their job isn't to configure Luns or configure servers It's it's to really make sure that you know, you're getting maximum IT load in the smallest footprint and lowest power But the new group that's out there is once you move to this virtualized model You know the people who are building the information services the ones who actually interact with the business units and say, you know We're gonna deliver you, you know, whether it's whether it's a collaboration service or whether it's a you know Decision-support application or whether it's a new service to your partners You know, that's the team that's gonna have more and more power more and more influence about IT choices And I think that's that's sort of the new development, you know the wild card in this a bit again is back to those You know the people who are suddenly gonna have to figure out then how do we deliver these all to smartphones and iPads? I can't tell you, you know talking to enough CIOs in the last year is saying, you know We get we get convergence. We get virtualization IT efficiency. We got to do all those things We have a roadmap, you know, we've got pieces to work out What worries me is that five years from now, you know, my primary user interface for all my apps has to be these mobile systems I don't have a clue how to do the mobile enterprise little data. Absolutely More big data but delivered in, you know, applets in the app store. So right that's that's right Or even gestural user data that's coming from a device. You guys must have been at VM world, right? I missed it this year. Yeah. Okay. Well one of the big themes, of course, you've heard this is VDI the virtual desktop infrastructure And that's whole thing shifting to the device. It's not even VDI anymore. That's that's sort of laptop and that's yeah Right. It's really it's that's legacy. It's it's, you know, it's it's your access manager Whatever that device may be whatever interface it's it's a lot more. It's bigger than virtual desktops, right? Right and you really heard that in VMWare's messaging, right? They really didn't use the whole VDI That was not as commonly mentioned. The whole ecosystem did because like VMware is always you know ahead of the marketing, right? And so I think you're starting to see that that shift and it needs to right because let's face it I'm the only guy who carries around a laptop We should do a tally of how many iPads in the HDS senior executives and yeah, that's going how about flash Is that is that change everything like everybody says it does or is that just the latest latest fad? From both the hardware and software standpoint Well, there's multiple layers here. So I mean clearly Bringing solid-state drives in into the drive frame in the storage systems is something You know it was started a couple years ago, but that was mostly to fix point problems I mean it's really just now with the transition to SAS architectures and other pieces and the advent of automated Data tiering that you can actually take advantage of that I do think that some ways though the more interesting part of flash right now is not in the classic storage system But it's showing up again closer to the servers or even closer to the VDI devices like a fusion IO approach Well fusion IO, but also you know things like what is it? Fast-cash I always get the multiple systems out there or even what what Oracle is so what's fast-cash? I think that's the net app But if you really think about the system level yes, well They're they're caching for specific types of data and one of the classic things that we ran into in VDI and also in big data centers with Virtual servers is the boot problem is DR If you know that that one time the whole system goes down You know how much time is going to take me to reload 10,000 virtual machine images? And so or now let's do it every day at 9 o'clock when everybody logs on so they're starting to say let's take You know, it's a highly deduplicate highly, you know replicated data set Bottle only 98% the same data for every single one, you know that shouldn't just be let's not put that in the standard Story system let's put out a cache that's really good at managing those at those image assets And there where you get a lot faster response time and performance on those kind of loads I think you're seeing that show up in analytics apps Seeing that show up in other places so this use of it's not tier zero and it's not main memory It's sort of this scratch capacity in the middle of your converged pool is the next interesting development You just have to make sure that that data does get staged to the storage at the end of the day Is that a big enabler or driver Laura on the storage software side? Is that what's driving a lot of the tiering or is it just really neat the need for generalized efficiency initiatives? Well, I think we're seeing it, you know being driven by Isolated areas of performance improvement And we have started to see Some some slow material impacts on the storage software front in particular, you know and as an example You know emcee has done very well with some of their fast V2 Capabilities and very well it's awful a small base, right? So it's still got a ways to go That's the sublun stuff right sublun migration. Yeah, and and you know, I think that's Important and the ability to do this migration between tiers is important But it really has to be in the context of some content some information And so bringing it back to I think Katachi's sort of vision an objective is really to manage Manage infrastructure given information policies, and I think that's ultimately where this needs to go is is not just management And migration of data at a physical level, you know LUN sublun page record level but but based upon a business role and a business policy And you know we're obviously a ways away from that, but but I think that's ultimately where it gets really interesting from a software So data classification policy engines Jack talked a little bit about that at least conceptually But it sounds like it's got a long long way to go in terms of upside potential I think so well you because you have the chance there is otherwise You got to go back and tell everybody to redevelop their apps a whole new architecture, and that's just not gonna happen that much For some cases yes, but for most part. Yeah, there's got to be a big value prop for them to do that And some use cases that there may be there if you want to develop in Ruby own rails and Hadoop go ahead But you know that's not exactly the traditional application base out there I think the value prop is in the data mining and data analytics space being able to do something with this Information that's just growing stale or over time and the value proposition in addition to the data mining and data analytics side Where you can actually mine the data and reuse it for some business value is also risk mitigation So managing information according to compliance rules, and that's always been a big theme with CIOs, right? I mean CIOs think about risk first and then then it's okay. Where's the business value and how do I pay for it? Yeah, I mean 10 years from now everything that's done at the infrastructure level is going to be time-stamped documented monitored And there's going to be a chain of custody around and everything that happens in the infrastructure Yes, you guys produce that study of really John Gantz does that digital universe study I don't know who's going to do it next year John as I understand semi retired is that right and that's quite a quite an undertaking and Do we have data do you guys have data on how fast storage is growing? Is that something like a fun fact that you have at your fingertips? I'm always hearing people quote you on that and that and it's all over the place Yeah, it's got to be a hard thing to track But you guys have the there's more visibility of the factory and the demand you do a lot of surveys Well, I mean we do keep track of the disk drives going out the door Yeah, right something we have a lot of presence on and we do track even you know the disk drives going into Enterprise systems versus going into your PC. So how should we be looking at data growth? Well, that's a that's a different question because you know we just talked about de-duplication and we've talked about thin provisioning in some ways You know data is growing faster than the storage by far Even inside enterprises. It's it's more about again if it's the same data 20 times Let's use something so I don't need so much storage. I think that's what people have woken up to is is You know, you have to be smarter about the storage load in it now having said that, you know You know, there's still a lot of inefficiency in it So, I mean, we're seeing that on average companies capacity growth is goes up is around 50% It was down obviously last year, but it was still over 33% on a per terabyte basis And you know this year is coming back strong and looks like it'll be well easily in that base Over 50 maybe even over 60 the problem we run into and this is for us is the issue is that that's a great measure If you're tracking x-rays or or you know MP3 downloads But if what you're really trying to track is you know those transactions or those data analytic things It's people actually don't need the capacity. Yeah, what they need are the IOs and in the past The way you got IOs was you bought a lot more disc and you short-stroked them and you were wasting 90% of the load So with solid-state disc, you know you're looking at IO and and I think this is one of the challenges for a lot of You know I see IOs and storage admins is we actually have to figure out a way to stop just tracking Dollar per gig for evaluating products. We have to find a better way to do dollar per IO So it's for a lot of the use cases It's really about the IO that you know that the amount of data the new data Those systems generate may is only is only grown at 10 15% a year But I'm growing the capacity 50% because that's the only way I can get the IO that can't continue And that changes the dialogue between the storage administrators and the app admin, right? It's not about capacity. It's about what are your IO requirements and it begins to become much more of an SLA driven discussion Excellent. We're here at the Cube live at the Hitachi information forum with Laura Dubois and Rick Villars of IDC Talking about data growth and the the Hitachi announcement and your guys perspective on the industry anything anything hot that you Want to want to tell the audience about you got any new studies you're working on or Any stuff you want to sell and now's your chance, you know, let's pump you guys up, right? You're doing a great job out there IDC number one market data and what do you guys how many analysts you up to now? Do you have a fun fact there? I know what within the company worldwide? Yeah worldwide thousand A lot of brainpower so well guys. Thanks for coming on the Cube and and sharing your insights with our audience It was great to have you great. Thanks so much. All right. Thanks