 Okay, we're back live. This is Dave Vellante and I'm with Wikibon.org and this is SiliconANGLE.tv's coverage of the Dell Storage Forum. This is our second year at the Dell Storage Forum and we'd like to bring you the top executives. We had Darren Thomas earlier. We have a number of customers on. The University of Wisconsin was on. And we also love to bring in the analyst community to bring an independent perspective on what we're hearing at the shows. And we're here with George Crump who is the founder of Storage Switzerland. Storage Switzerland is a consultancy. George is an analyst, a blogger, a very prolific writer. George, welcome to theCUBE. Hey, thanks for having me, Dave. Appreciate it. It was really a pleasure. I appreciate you taking the time here. We're at the Dell Storage Forum. Now were you at the Orlando event last year? Yeah, sure was. You were, okay, so the second year, I feel like it's a step up from last year. I think they're moving in the right direction. Yeah, absolutely, I think so. I think the message that I'm slowly getting from Dell is, hey, we're really serious about storage. Last year was sort of, okay, yeah, you guys are maybe paying attention to storage, starting to get it now. They're serious, right? Yeah, I agree. Dell is clearly relevant in storage. There's no question about it. And of course, they're competing with a lot of larger companies. We're going to talk about that. You have a good line of sight on what's happening in the competitive world. But the transformation to Dell has been very interesting to watch. I mean, Dell basically was a reseller of other people's technologies. We all know that Scott McNeely in his unique way one time said to me, Dell's idea of innovation is the new form of W on the keyboard. Tell Michael Dell I said that. He would always bust balls around that. But Dell is really geared up to bring its own IP. And of course, the message they're giving to us and the customers is, it's customers first that we're doing it because that old legacy stuff, euphemism for Clarion, is outdated and hard to use. That's why we went out and bought these new assets. But the reality is, there's another side to that coin which is the impact that it has on Dell's business. And Michael Dell is architecting the transformation of Dell. It's a higher margin business, obviously, in storage. And Dell could be dangerous, don't you think? I mean, they're used to living in PC margins and now they're playing in storage. It's like heaven for them. Well, I think there's two things. I think there's clearly that. And I think also because they've been so focused on that mid-market and they really pay attention to simplification and making things easy. Well, last time I checked enterprises don't mind easy stuff either, right? And so that, I think, is going to play well in the enterprise, right? What we see a lot of times is enterprise guys buying mid-market solutions, specifically because it's more cost effective or easier to use or whatever the case may be. A lot of companies, I feel like we're forced into easy. Making things easier. And some of them arguably bolted on easy. Dell had the advantage, I guess, of starting with a blank sheet of paper and it could acquire companies like Equalogic and Compellent whose raison d'etre was to simplify IT. Right, I mean, both those guys, right? Their whole go-to-market was easy, easy, easy, right? I mean, I was convinced when the, I remember when the Equalogic guys first started going to market and the four sales guys would walk around the Equalogic box and plug it in at customer sites. I mean, and it worked. And of course, back then, if you could plug in a saner and it actually worked, that was a sale, right? It's harder to make easy happen when you got, obviously, a big legacy in Stallbase but what a lot of companies have done is you take, for example, IBM, they went out and bought XIV. Right. HP buys 3-par, EMC buys Isilon. I guess that's not, wasn't so much of an easy play but for instance, they develop a VNXE. So they just create a whole new product line. So it seems to me that you got four or five companies that are really viable in this space. Dell's obviously one of them but what do you make of the storage landscape these days? I think that's a good way to really sum it up. I think what we see is what I like to look at is how people are doing with acquisitions because for these bigger companies, their growth is going to always come around or a lot of it's going to come from acquisitions, right? And Dell, and to a large extent, EMC, both really execute well on buying two types of companies. Dell buys pretty mature companies and in that area, you just kind of don't want to break it, you just want to put it into your existing engine and accelerate growth and maybe later on do some integration work, things like that. Compellent, you're saying would be an example of that. Yeah, compelling, ecologic, clearly, right? And then I think there's the guys that I call you have to incubate some and that takes patience and interestingly, I think we've really seen that with Dell, right? With the Exynet technology and the Ocarina technology. I mean, I'm sure I think I would throw more into the more mature area but those are good acquisitions that I think they've made. Yeah, and I guess you're right. I mean, EMC's been a good, solid acquirer of companies. I mean, actually great acquirer of companies that you can throw in VMware into that mix and some of the others in the storage business have mixed track records. Why do you think Dell is so good at it? You know, I'll tell you, this is totally off, this is no analyst's end point but I think that because Michael Dell is still so heavily involved in the mix that it's still very much his company, right? He's still the namesake and everything where at other companies, it's more classically other people's money, right? And so they don't, I don't know if they've put as much thought into it, it's not as personal as it may be to Dell. My sense is, and they certainly would never, I don't confirm this, but my sense is it's just a more personal acquisition for Dell. They take it more of a family sort of deal. You're kind of including somebody in your family. You want to make sure you're, you know, you don't want to invite them to barbecue and stuff like that. There's names in the box too, right? Yeah, yeah, I mean that's a big thing, right? That's a good point. You know, it's funny, I mean I've watched and been around for a while and you've seen transitions in the industry. I mean, certainly remember, as an East Coast person, we're here in Boston, remember the old mini computer days with Deck and Prime and Wang and DG and those companies are Apollo, they're gone, right? And they were absorbed or near a brink of death and you know, somewhere at the Chapter 11, whatever. It seems to me, George, that the PC era companies, Microsoft, certainly Dell, the guys who really made it big during that age, it's like their executives saw the fact that they crushed some of these larger, really well-established mini computer and other companies. And I think they're much more proactive and smarter about survival. Right, you know Dell is a company with, I think $14 billion of cash in the balance sheet, they've really effectively, I remember DG in the day, they had a ton of cash in the bank, they owned a bunch of real estate, they didn't know what to do, right? They were just like deer in the headlights. Dell is dramatically different. And I would say Microsoft too, how many years have people been gunning at Microsoft? Right, yeah. So, what do you think is up with that? Is it a sort of pride? Have they seen the play before? Are they just smarter? Well, I don't know, I think that because history is happening in closer cycles, you're more attuned to what happened. I mean, most of the guys, or a lot of the guys in the Dell executive team have been around in that mini computer era and they saw that happen, right? And so I think you're right, they can see the handwriting on the wall that the PC is not the long-term viable market that it used to be, right? And I think the transition now to a bigger, you know, more focused sort of storage and server, I think Dell understands that, and I think some of the other guys do too. Yeah, they do, and they've got, many of them have strong VC arms, right? I mean, I saw EMC by ExtremeIO, they had an investment in the company. Dell today just announced Dell Ventures, I believe it was Dell Ventures, they announced an investment in a company in Ann Arbor, Michigan that basically made highly efficient motors for scooters, which was really amazing, you know, seeing that type of, that's talking about a, not even an adjacent market, a total different play, but very highly innovative move. So you're seeing Dell, and they've tried some other things, like Dell TVs and other consumer electronics, didn't work out so great, so I'm impressed with these companies that are putting their telescope on and really trying to get ahead of the curve. So let's talk storage. What are the big mega trends that storage Switzerland is following in storage, and what are some of your favorite things that you're watching? Well, clearly flash, right? SSD flash is big. Maybe not so much a trend, the other trend that we're seeing, just from analytics off of our site and things like that is the basics, right? There are times where we just sort of assume that everybody knows what an IOP is and everybody knows what bandwidth is, and you know, the number three ranked article on our site right now is what are IOPs? Well, what does that mean? Well, it probably means we've got a whole bunch of people graduating and they're hearing this term IOPs, and they're, what does that mean? And they're probably Googling that term exactly, right? So I see two trends, we've got a whole new group of IT people coming into the market that they're just hearing these terms and they're just being forced to assume that they should know what they mean, right? And then the more kind of elaborate stuff like flash, like cloud, you know, those sort of things are all becoming pretty big. So that ladder, that what is IOPs? So Google, what is IOPs? See where storage Switzerland's article comes up, it'll be interesting to see. But so that ladder trend, I think bodes really well for Dell, which is really targeted at the IT generalist, right? Right, absolutely. Yeah, and you know, I was on a, I probably shouldn't say who, but I was on a briefing call and they were announcing a new PCIE card that could do a million IOPs a second, and I didn't say anything. And there was this sort of long, uncomfortable pause and the guy goes, well, isn't that impressive? I thought, yes, sure it is, it's impressive. But nobody knows what that is. What does that mean to the guy? Does that mean he can go home at four o'clock instead of five o'clock now? What does that mean? Does it make more money for his company? Yeah, they don't know what a million, and can you, could you really take advantage of a million IOPs even if you had it, right? It'd be like if we increase the speed limit to 200 miles an hour, well, if the cart off all the traffic's going 20 miles an hour because it's rush hour, it doesn't really matter, right? Don't you think that, I'm going to make a statement, tell me if you agree, disagree, that historically the storage industry has done a really crappy job of selling business value. They've done a good job of selling to storage admins, features and functions, but not really appealing to the business advocate, do you agree with that? Oh, 100%, in fact, I think that is the key reason why you see so many flash startups instead of, the manufacturers clearly have all brought flash to market, but they've all kind of hinted that it's not going as well as they thought it was, but the flash startups are doing pretty good and because those guys understand that they need to communicate business value where the more traditional manufacturer is really still in that sort of IOP, speeds, feeds, you know, those sort of things, right? And when you're selling something that's 15 times more expensive, you got to be able to explain to somebody why you'd want to pay that, right? Yeah, so let's talk about that a little bit, and it seems to me that flash has the potential to really have huge business impacts, productivity, you know, if you measure even something as revenue per employee or something like that, it seems like flash and the development of new applications really could change a lot of the way in which companies transact business. So you're seeing bits and pieces of that, I guess today with like a Fusion IO and you see the activities they have at Facebook and Apple and maybe some of the other, you know, database-intensive applications. My question to you is, we saw as part of the recent history of storage virtualization with Equalogic and Left Hand and Component and 3PAR and Isilon and I guess I threw a data domain in that mix, one of the greatest wealth creations in storage that I've ever seen. Do you think flash will be a repeat of that? Do you think it'll be bigger than that? I think it'll be bigger. I think that most of those storage companies are going to have to look at their existing portfolio and say, we can't do it like we used to. It's going to require different types of architectures and things like that. And so yeah, I think that these other companies that are sort of the next wave, if you will, right? I think that there's going to be some massive wealth creation. People are very, the people being the traditional storage companies are very conservative or careful about insensitive to the notion that flash will replace spinning disk. There is a line of thinking though that basically says that all active data will be on flash. Do you believe that? Yeah, in fact, I made a proposal once to do a presentation that in five years there'll be two types of storage, flash and tape, right? And you would never use those two in a sentence, but if you kind of map out the long-term coordinates on both of those technologies, they're the only things that can get faster, right? And hard disk has sort of stuck where it is, right? Now, there'll still be some hard disk, but you could see a world that is mostly flash, a little bit of hard disk, and then a lot of tape. Yeah, so I say, there's a great sensitivity of the vendor community when you'd say that, that they don't know, there's a place and they sort of craft their messaging, but it's just if all active data is going to be on flash, well, where's active data live today? It lives on spinning disk, so. Right, well, but I think that the funny thing is, that I think that technologies like what Dell talks about with their fluid data still are going to apply in flash, because I think we're going to have levels or tiers of memory-based storage, right? We're not going to, we'll have a simple example today would maybe be MLC, SLC, and 3-bit or TLC, but as we get into the next types of memory, we might have a mix of that plus the MLCs and SLCs. People say that people in the enterprise space say that if we had to pick a technology to essentially be a persistent technology that we would use in the enterprise, we never would have picked flash, because all the gymnastics that we have to go through, you know, to protect the data, and make sure you deal with wear leveling and make it reliable. At the same time, the reason why flash is where it is is not because somebody in the enterprise picked it, it's because of Apple. Right, absolutely. So my question is there's a lot of discussion about, there's some negative friction in the enterprise about flash because it's maybe not the most enterprise-ready technology, but people have dealt with that, and there's a lot of discussion about follow-on technologies, memory store from HP, for example, and others, you know, a lot of stuff in the labs. What's your take on that? Will those technologies be able to compete from an economic standpoint, or is it really going to be the consumer market that drives the market? Well, that's a tough call. I think that the investment and just the production around flash is so big that I don't know if anything else will be able to overcome it near-term, say within the next five to 10 years even. I just think there's just so much invested in there. I mean, clearly as we get smaller on lithographies and things like that, you know, the flash technology becomes increasingly challenging, but as you start to look at environments that are entirely flash-based, well, you know, EMC's acquisition and some of the other guys that are out there, that changes the game, because if you think about how we're using flash today, it's sort of the worst way to use it, where we're caching stuff, it's up there for a little bit, and then we're pulling it back off and we're putting more stuff on it, and so we have this huge turnover rate. Well, the last thing we really want to do is do a lot of writing to flash, right? Well, if it's all flash, then you don't have that turnover, right? So I think that as that use case evolves, we'll start to see better reliability. The other big thing, of course, is as we have enterprise-class systems protecting flash, if we have a failed flash module, it's not that big of a deal, it's no more a bigger problem than a flash hard drive. So if you take that sort of predominantly flash or, you know, all-flash scenario, and we've talked about this a lot, obviously, the importance of software. Right. To manage that whole hierarchy. Do you see it as a one-way trip to Bitbucket? In other words, we've got the data in flash, okay, and once we're done with it, it's going to go back to the Bitbucket and probably live there forever. Maybe we pull it in every now and then. Yeah, I would say that that's probably the typical scenario. I think on software, though, I think that the zero latency or the low latency of memory-based technology also changes software because now, you know, the software that we've had from all the major manufacturers really changes because they were counting on latency in the hard drive, right? I can remember when, and I think it was NetApp, really, we would talk about, well, we've got to figure out ways to use more of the storage controller processor because it's sitting there doing nothing most of the time. Well, you know, in a heavy flash environment, that's not going to be the case anymore, and so all of a sudden, the software becomes a bottleneck and so they're going to have to become more efficient at software code. Yeah, so we had Pauline Nist on theCUBE. We've had her on a couple of times. We had her on HP Discover, had her on Oracle OpenWorld a year or so ago, and she used the term the horrible storage stack and it stuck in my mind and what she's talking about is not just the spinning nature, but also, if you think of, you know, just even something like the SCSI protocol, the overheads associated with that, and I'm very much intrigued by what FusionIO has done with its VSL technology, being able to essentially do atomic rights. What's your take on that? Is it really a differentiator? Is it something that Intel is going to, you know, going to ultimately, you know, develop and compete very effectively with them? Are there other entrants there? Or does FusionIO really have a significant lead in your opinion? Well, I'll tell you, it depends on how well they can get the community, the development community to embrace that stack, right? If they do, I think they have a big advantage, right? The ISV community writing to their API. Hey, well, I'll tell you the other angle for it, that's interesting for like a FusionIO or even for some of the other guys is storage systems that then use them as part of their stuff, right? So instead of going out and buying yet another PCI thing, I mean, we've seen, I can think of two startups that have FusionIO embedded in their storage system as part of their overall strategy and they're leveraging that API set and they've customized their storage software to take advantage of that high PCIe transition. Are they stealth startups or are they like a data core? Well, specifically, I was thinking of Nutonix, which is a converge play, and then also NextGen, out of the box, right? Okay, so, and then so you've got the sort of memory extension thing going on and part of me doesn't even see that as a storage, you know? It's something new and different that could change, as you say, applications. And then you've got the other end of the spectrum like a violin memory, you know, which is, I've called them the data domain of flash, right? You just sort of plug it into the existing infrastructure. You get huge valuations. They've raised $160 million. What do you make of those guys? Again, I mean, they're on a good track. You know, their issue is going to be how well can they transition? They've partnered very well with HP. I think that's where they've seen a lot of attraction. Their thing is going to be how well can they come into a shareable, more of a true appliance type of nature without any latency and things like that. So those would be the key things. Anybody else that you're really, you're watching closely? I mean, I know you're watching everybody closely. Yeah, I know you've got to be Switzerland, about this. Well, in the flash space, again, I like the all flash plays. I think that they have an interesting, even if people don't use it in that nature just as a kind of a, to handle a specific problem within the environment, I see that. So guys like Pure Storage and Nimble would jump to mine immediately, solid fire again out of Colorado would jump to mine. So I think there's a lot of guys in that space that can do a lot. The other guys that are kind of interesting are the guys that are integrating like a cache or an SSD layer and then using sort of SAS or low end, nearline SAS on the back end. Just to drive, so that drives flash into the SMB space and they're really using it, not necessarily to have great performance on the SSD, they're using it to make slow drives go faster. What's Dell have to do, in your opinion, in flash to maintain leadership or grab leadership? Well, I think they have a lot of the foundational stuff. The RNA stuff could be big, right? And so they have the ability, I think in the enterprise class, this move to server side, pushing the data closer to the server just makes a lot of sense. It's who's going to have the best technology to execute there. I think they've got the foundational stuff to do it. I don't think we'll see it until next year, but I don't know if that's necessarily a big deal. I don't think everybody's going to be racing to server side flash tomorrow, right? So I think that's key. I think the other thing is going to be architecture. We talked about that earlier, is can they tweak the ecologic and compelling architecture is enough that they can get full SSD bandwidth out of that scenario. What do you think of the Extreme I.O. acquisition that reported 430 million, although Rich Napolitano said to me on theCUBE, it wasn't that high. So I don't know what the number was, but it was big. Right, you know, what EMC does a good job of is buying the right company at the right time. I mean, so you almost kind of can't argue with whatever they do from that space. Now it really comes down to how well they execute, because they've had cases where they've executed very, very well, and they've had executed, not so much, right? So to me, the challenge that Extreme I.O. brings to them is can they execute a high value type of product, right? It goes back to what we started with, right? Can you explain why I should pay 15 times more for this than the other thing, right? And if they can do that, I think they're in good shape. The difference, though, is where, when they bought ecologic, when they bought Iceland and Data Domain and those guys, there was a team of people that were able to help their regular sales force with that transition. They kind of left them alone in a way. Exactly, there's no team. I mean, there's a team in Extreme I.O. They don't have sales people all around the world, right? And then that's going to be a hard thing to do. Yeah, that all-flash array looks like it's going to take on the traditional tier one block-based storage space, which I was very intrigued that they put the Extreme I.O. acquisition under Rich Napolitano's group in the mid-range, the unified storage group versus Brian Gallagher in the Symmetrics Group and the Symmetrics and Virtualization Group. Why do you think that was? Do you think they're going to create some internal competition, or was it just a matter of management bandwidth? I don't know, because of all the pure-flash stuff that's out there, I think the Extreme I.O. could make a case to be a real enterprise solution. It had the scalability, the HA stuff. So I think it's more that they're still trying to figure out where that group fits in. And I think that the high-end enterprise group within EMC is still treated special, right? And so I think they would have to kind of earn their way into that. All right, George, we're getting the hook here. I really appreciate it. People want more information? Storage-Switzerland.com. Great, excellent resource. It's all open-free, no firewall. Really great to see you. Thanks very much for coming. Keep it right there. We'll be back with more from the Dell Storage Forum, live from Boston. This is theCUBE.