 virtual storage platform. I'm here with Ray Lucchese and David Flyer. Ray Lucchese is the president of Silverton Consulting. Consulting, Ray, welcome to the Cube. Thanks, thanks Dave, thanks for inviting me. You're welcome. And David Flyer, you're a Cube alum. Welcome back. Thank you very much. Good to have you guys. So Ray, Ray Lucchese, very astute analyst, former STK, you know, in the heyday, right? You guys were doing some really innovative things. Literally, heyday. Yeah, making, you were making a lot of hey. Hey, Rhymes with Ray. So rayonstorage.com is the blog. That's correct. That's a blog. I'm on Twitter, Ray Lucchese. Very prolific blogger. And of course, you guys know David Flyer of Wikibon.org. So guys, big event here. This is very, very non-Hitachi-like, all this great marketing. Ray, what do you think? I think the marketing push is excellent. I think it's been something they've needed for a long time. I think as they move from, you know, a more OEM kind of environment to a more end user perspective, and they've been there all along, but this is even more of a push they need to have much more substantial marketing. And this is great. Yeah, and you're alluding to a couple of the deals, right? The Sun relationship, and of course, they've got an OEM relationship with HP, and we'll talk more about those. So David Flyer, let me ask you, what are the high points of this announcement from your perspective? So the high points of the announcement is the extended architecture of the new box. And it's taking a very different approach to others in this marketplace. It's creating a box which is flexible. It is a very, very powerful box. You can adapt it to work in different situations. And the virtualization strategy, what they've called Scale Deep, they've extended that. And it offers a very clear, different strategy to both the MC and the IBM, especially in the tier one. Anything you'd add to that, Ray? No, I think it's correct. I think, obviously, the USB has been focused on storage virtualization for quite a while, and the USB elongates it or extends that advantage. I think it's definitely orthogonal to what EMC has produced from a VMAX perspective, IBM's DS8000, and some of the other competition out there. So from that perspective, I think there's a lot of discussion that you could go around with respect to VMAX versus VSP from an architectural perspective. Well, we all three of us talked to a lot of customers. And my opinion is that in terms of the true high end, the real mission critical types of applications, it's really EMC and Hitachi are the two gold standards. IBM has the DS8000. In my opinion, they haven't invested nearly the way that EMC and Hitachi have. I don't know if you guys agree with that or disagree with that. Well, yeah, they certainly have done something with respect to the functionality to keep updating the drive hardware and some of the other functionality. Easy tier is nothing to sneeze at. It's a very nice dynamic tiering capability, but they certainly haven't restructured, re-architectured like their competition EMC and Hitachi. You can't leave out NetApp and you certainly can't leave out some of the other players. Like 3par? Yeah. You know, 3par hates it when I don't include them in this discussion, but it's going to take them time to mature in my opinion. David, is NetApp there yet in your opinion at this level? To be tier one, you need several things. You need the architecture. You need the ability to focus 100% performance on a particular application. Both NetApp and 3par tend to be, everything is okay. Good enough performance. So this VMAX versus sort of VSP is only part of the story, but I don't want to leave it quite. To me, it's like Republicans and Democrats. They both bash in it. They both solve problems for their constituents, you know, but really taking a different approach. I would describe, and I wonder if you guys are, you know, the technologists, describe this sort of VMAX as more of a scale-out modular and the VSP is sort of this tailored, really purpose-built for the problem. And they're talking about scale-up, scale-out and scale-deep, which has never been done before in the industry. So I guess my question is how much of that is real and how much of that is vision? Do you have a sense of that yet, Ray? Have you been able to analyze it? So I think that, you know, they've obviously had, you know, I'll call scale-deep kinds of capabilities in the product and the USPV and the prior version storage virtualization has been there. They're there. There's no doubt about it from that perspective. Scale-up, you know, adding, you know, the small form factor disk drives, the two-and-a-half-inch disk drives is a key deliverable from a density perspective and performance perspective from a back-end storage. Scale-out is an interesting discussion. And that's where the nut of this whole VMAX, VSP, 3-par, NetApp kinds of things starts to hit the road. You know, obviously VMAX is a more commodity type of environment. You know, you can use a VMAX. You can put software on it. It becomes, you know, Vplex. You can put other software on it. It becomes something else. It's something that can provide a lot of different functionality on that same box. And if you look at the USPV or the VSP, for that matter, the VSP is a storage virtualization engine. Everything that revolves around storage virtualization, they can do. And, you know, Vplex is a storage virtualization engine versus VMAX, which is just a storage. So it's kind of, it's interesting to see how the two products are differentiating themselves. So, David, Floyer, Vplex is interesting. I was saying with John Webster earlier, is Vplex is sort of a confirmation that Hitachi's and IBM's SVC's virtualization strategies have worked. Would you agree with that, David? Oh, I think so. Absolutely. And Vplex is trying to differentiate, but it's new. It's just, you know, getting started. It's new. It's getting started. And, of course, it's got to integrate with the VMAX itself. So there's a whole integration issue that EMC have in bringing that together. So I'm sure they'll be successful. We have ESG on later. And ESG published a study that showed it was basically from the U.S. and, David, you picked up on this, is that I, that EMC was number one in VMware penetration. Hitachi would be, if not number one, close to number one in storage virtualization. Would you guys agree? In tier one, definitely. Oh, absolutely. Yeah. Okay. Yeah. All right. So one of the surprising statistics, and let me throw this out, is that, and I think Brian mentioned it, you know, roughly 50% of the subsystems use storage virtualization. That is higher than anything I would have expected to see. I mean, in the past, it was, you know, 10, 20 percent tile kinds of thing. 50% storage virtualization. So what's happening? What's driving storage virtualization besides the M&A and the client acquisitions and stuff like that is a key mystery in my mind. And so, and so, you know, VSP and USP have kind of been on the forefront. SVC is there. Vplex is coming online, and NetApp's got their own virtualization engine. So you can see all those players playing out the virtualization card. Is that drive to simplicity maybe? It's a drive to simplicity, but again, at the high end, one of the things, one of the big costs is the cost of migration. It's $50,000, $100,000, and additional costs on it. Technical upgrades for data center movement. All of those issues about migration really add to cost. And if you take that strategy and you can eliminate a high percentage of that cost, that's real money and real risk reduction. One of the things I was impressed with this morning was Jack Domey's keynote, where usually Hitachi keynotes are very product-oriented. Now, Hew, I think did a great job of describing the product. Jack really, you know, didn't really touch on that. I mean, touched a little bit on it, but he really set forth a vision. And he talked a lot about this content cloud and really bringing together unstructured and structured data, in my words, not his. He was using that term, the content cloud. He talked about governance. You don't usually associate that type of content cloud with these types of mission critical systems, do you, Ray? No, it's an interesting vision, and that's what I call it at the moment. I think as you integrate content, file, block, and other types of data sitting on a storage computer, as Hew said, it becomes a very interesting play. Here's the question. Does that sort of storage and information and data consolidation still make sense, or do you want to have different appliances do that sort of stuff? Data leak protection, governance here, centera for content? What's the right answer to that? I'm not sure anybody has the right answer. They're all valid and viable approaches. Well, it definitely demonstrates innovation on Hitachi's part, though, doesn't it? I guess it all comes down to cost, doesn't it, David? Yes, absolutely right. Their vision is they should manage it. They're not saying that everything should run necessarily on the platform, but they should manage it from their visualization point of view, manage across it. That, again, is a very different vision. Can you attach a centera to a USB-V? I don't believe so, but that would be a question for you. Yeah, well, I guess, yeah, we'll ask, but I mean, I guess my point is maybe you can't attach a centera, but you can attach so many other devices that that, to me, is the real power of that vision. But it's not unclear that you couldn't attach HCAP to VSP or the HNAS solution to VSP. All those things can be on top of a VSP block structure environment. Is that the content cloud? Is that what he's talking about? It doesn't feel like it. Some archivist stuff, some Bluark stuff and the management framework. Yeah, the management framework, I think, is the number one. On top of all that. On top of all that. So that you can move stuff from production to archiving to wherever it is. Oh, that's interesting. Actually have some sort of life cycle. Would you guys like to see HDS or Hitachi take a more active role in M&A and maybe go after a company like a Bluark? What do you think about something like that? So Bluark is a natural acquisition. If they want to play that, they've been doing good work and good business in my mind with Bluark for a while there. And if they continue that relationship, does an acquisition make sense? Certainly. I mean, to some extent, Bluark sitting there with a hardware accelerated file system. And Hitachi, if nothing more, is taking the storage computer and building ASICs and architecture and throughput in the 5th generation crossbar. It's a hardware architected soft storage system. So the merger of those two makes a lot of sense. Yeah, I would think that that would really send a ripple effect through the industry. And to me, underscore the differentiation that we've been talking about. I mean, I should invest in Bluark. I don't know. We've made some good predictions here in the Cube and sometimes we're not always right. But Hitachi is not prone to those types of major acquisitions. I mean, it did archivist, but it's been relatively reluctant. So you hear a lot about the Hitachi transformation. I think it would be a great play. I think it would be a good play. And I think it's Hitachi should become more aggressive in, especially with software technologies and making. You know, these guys are spending $7 billion in R&D over the next couple of years. I mean, I don't know, billion here, billion there, you can actually buy something. How about, well, Hitachi is a $97 billion company. Revenue, yeah, yeah. It was pretty impressive. How about VMware? What do you guys think of the VMware story that you're hearing today, Ray? So VMware is driving server consolidation, data center consolidation. So as these customers get more and more virtualized, they need more and more enterprise class storage, reliability, availability, and support 24 by 7 in order to keep their environment up, it's driving enterprise storage as if nothing like it has been in the past. Yeah. And their implementation of this lovely acronym, wasn't it? Yeah, yeah, yeah, yeah, yeah, yeah, yeah, yeah, yeah, yeah, yeah. VMware needs some new native directors. It's gonna have a V in front of it. Their implementation is, has looked very, very good and very complete. Hitachi's implementation of this. And it's very early to market. I must admit, they're not the first, but they're really early, aren't they? We're talking about this year, right around the time, I think that November timeframe, they're talking about some of the benefits that are coming from this, in particular, the very technical SCSI, there's some SCSI locking, which goes on and they are getting rid of that by pushing it down to the array. We just lost a thousand viewers. Yes, but your point is it relates to clustering and scale out and my point is that that is a constraint on the number of VMware machines you could run on a server. So by eliminating that, they're actually able to increase the number of virtual servers. So it's a scalability thing. It's allowing VMware to scale the more and more virtual machines by providing hardware-oriented facilities that just make it perform a lot better than without it. So that's part of the scale-out vision, right? Potentially. If you take the scale-out from the storage perspective to the application and virtual server perspective, yes. So scale deep, scale up, scale out on a scale of 1 to 10, 10 being, you know, with air, one being, need quite a bit of work to do. What would you guys say for scale deep for Hitachi? Oh, there are nine and a half. Nine and a half, David, do you agree with that? How about scale up? 10.5, in fact. To me, they're the best, in scale-out. Scale-out is... Scale-out is, in my mind, storage capacity. So they've got this, you know, with deep, which is their card, right? 247 petabytes of storage behind this. It's kind of hard to beat that. Okay, and then scale up, you guys would say, or we're hearing disagreements. Scale-out. How about scale-out? I think we're okay from an external storage perspective. Internal storage, they're, you know, both the EMC and they can support 2,000 disk drives. I mean, it's... But I think from a raw performance point of view, the architecture, the ease of the enable- It'd be interesting to see the performance comparisons between a VMAX configuration and a VSB configuration. And, you know, we'll probably won't see that except in a few esoteric applications. Well, that's under the MPC record, right? And... Yeah, at the moment, I guess I'd have to look at it. I would think this would beat it, but the EMC doesn't participate in those. But now, scale-out. What about scale-out? Scale-out 1 to 10, where are we at? I'm at 30 current. That's the question. If you look at something like VMAX, it's not apparent that it couldn't go from 4 to 8 to 16 to 24 of these VMAX nodes without breaking a sweat, without having to change hardware, without having to change back in, without having to do anything. These guys, they've got a blade cabinet. So you've got one control chassis, you can go to two control chassis. Can you go to three? Can they go to four? I doubt it. It's not apparent. Okay, so the scale-out would need some work. The other two, two out of three is not bad. I mean, still, the vision, this is unprecedented, right? Scale-out, scale-up, scale-deep. And... They would achieve the scale-out with the scale-deep. They would argue. Yeah. And perhaps it's performance, right? By adding more and more external storage, you can increase the performance of the VSB. Up to some limit, whatever that limit is. Okay, so they can tick that scale-out box and... Well, they've got the scale-out. They go for two processors, the four to six to eight. You know, they've got a certain amount of scale-out in the VSB. It's no doubt that it's there. They have the advantage of it that it's integrated, whereas V-Plex and V-Max still need to do it. And advantage or disadvantage of integration. Disadvantage of integration is that the cost of doing that technology, that A6, the fifth generation of cross-bar, the sixth generation of cross-bar, keep all these things up, is a challenge that they have to maintain. It's tailored, it's custom. It is custom, but it's, you know, $7 billion in R&D. That's where that goes. It's a $97 billion company. Yeah. Maybe they can afford it. All right, we're here with Ray Lucchese and David Floyer. Ray Lucchese of Silverton Consulting. rayonstorage.com is the blog at Ray Lucchese on Twitter. At DFloyer, wikibon.org. Guys, thanks very much for coming on and having a great session on theCUBE. Appreciate it. Thanks, Dave, thanks for coming.