 Live from Las Vegas, it's theCUBE. Covering Edge 2016, brought to you by IBM. Now, here are your hosts, Dave Vellante and Stu Miniman. Welcome back to IBM Edge, everybody. Hashtag IBM Edge, go to ibmgo.com. This is theCUBE, the worldwide leader in live tech coverage. Our friend Eric Kurtz-August here. He's the vice president of product marketing for the IBM Storage Division and Software Defined Division. Great to see you again. Thanks for coming up. Great, well thank you very much. Always love coming to theCUBE. You're looking good. No Hawaiian shirt? You must have been presenting or something. I was presenting this morning, so I couldn't wear my normal Hawaiian shirt, but next cube I'll have my Hawaiian shirt on. No tie, obviously. No, I don't think I've worn a tie. You're old one? I think I own one or two, but I don't think I've worn a tie for a long time. Dave's got the tie for us, so, you know. I got it covered. So, you look good. How was VMworld? VMworld was incredible. You know, it was here in Vegas instead of the normal place San Francisco, so some of the booth layout wasn't as good, but there was a lot of energy around it. Obviously the EMC, Dell, you know, weren't going to do the acquisition next Wednesday, it was announced on Wednesday, so there's a lot of hubbub around that. We see it as a great opportunity for us. VMware has a partnership with IBM's cloud division, so they did a big to-do about that. The block was there. It was huge. The block was on stage with Pat. On Monday, so, we had a much bigger booth than normal, so it went very, very well, and we have a bunch of products that are, you know, tightly integrated with VMware, Vasa, VAI, vCenter Ops, you know, whether it be the store-wise products, the XIV, the A9000 flash system, so, you know, for us, it was a very strong event. Very strong place. I mean, you know obviously a lot about VMware and the VMware ecosystem. You were at EMC for a number of years. You guys were very good at leveraging that relationship. Helps when you own 80% of it. But what's your take on VMware as, you know, in a state, if you will? Everybody, you know, sort of the tongue-in-cheek was VMware's the legacy. Huge, huge business. What's your take on VMware as, you know, from a customer standpoint and the opportunities for IBM? Well, from our perspective, you know, we're working with any hypervisor. So, VMware, Hyper-V, KVM, OVM, as you know, containers are the hot thing, so we're working with container partners as well. So, however they want to put the applications, whether that be in a container configuration or some sort of virtual machine, of which there's four choices right off the bat, we want to make sure that our customers are always taken care of from that perspective. I think the biggest issue from a VMware perspective is where do they go now that they're part of Dell? You know, when I was at EMC, Joe Tucci had a very clear way, which is they were the Switzerland storage and the Switzerland servers. And you guys at EMC don't complain that they're working with IBM, don't complain they're working with other storage vendors because their job is to make money. And the way to do that is for them to be Switzerland. Michael Dell's personality is not Joe Tucci, so it'd be interesting to see in the long run, how does he treat that business? Does he treat it the way Joe did, which is they're the Switzerland of, so they'll be able to work with us, no problem, and other storage vendors, same thing on the server side. So, it'd be interesting to see how that really plays out because that was quite honestly EMC's magic with VMware was kind of leaving them alone and letting them do their own thing as long as they made the right revenue targets and profit targets and they did. And I'm not sure that Dell has that same sort of temperament that Joe Tucci did. And there's some signs that that's shifting a little bit. Nutanix is a good example. Now, when Chuck Hollister was there, he was blogging and sort of attacking Nutanix, the whole V-SAN thing. Some of that's Pat, right? Pat has more of a vertically integrated, if I can use that phrase, you know, strategy. And some of that could be Michael, but are you seeing indications that you're not sort of an equal partner as you used to be in the VMware ecosystem? So far not. So I'm actually the storage executive sponsor to VMware and used to work for Pat Gelsinger, who's the CEO. You know the lay of the land. I know the lay of the land. And so I'm the exact sponsor. To read the tea leaves. And so far, you know, all of the things I was at the VMware had their partner exchange conference, which was in April and they had all of the big OEMs in and Michael addressed everything and says, don't worry about it. They're staying independent. I expect them to work with all you guys, just like they always have when it was EMC. We sounded legitimate. So the proof will be in what happens in the future. Don't you think that would be a huge mistake though for VMware to sub-optimize its business in order to help, you know, the Dell business? Wouldn't that be a strategic mistake? So I would agree that would be a strategic mistake. There's a lot of options. I view it as there's sort of three options. A, he sucks them in. And so now everybody runs to OVM, KVM, to Microsoft for Hyper-V and to the container, all the container startups. So they run away over there and that could hurt them in the long run. The other option is that he keeps it, keeps owning as much as he owns, but takes the Joe Tucci attitude. The third option is truly to spin them out completely. Say Pat, go make your thing, make the money. As you know, if you calculate out what a spinout would generate from a Dell perspective, it would pay for the EMC acquisition and probably have a couple billion left over. So I don't know if Michael's going to do that. And then, of course, the other option is to do a partial spin, like what EMC did one time, and you know, I think you might have still been at EMC when they went down to 19.9%. They still were the biggest shareholder, but they're not legally now the owner anymore. So they could do that and that would free up a bunch of, generate a bunch of cash that they'd used to not pay off the entire acquisition of EMC, but a partial acquisition of EMC. From our perspective, obviously, if they stay independent, that's great. If they don't stay independent, that's okay because we're already working with Microsoft with Hyper-V. We already work with container people. We already work with Oracle Virtualization Manager, and we work with, of course, KVM as well. So we have always been that way and a good storage vendor wants to make sure that whatever sort of virtualization layer, or now I'll call it a container layer, that end user chooses that you want to be able to do. You know, a lot of big, big enterprises, even though that's a stronghold of VMware, a lot of those guys have definitely Hyper-V because it's usually bundled in for free in their enterprise license agreements, or some of them obviously because of the KVM being basically free, not really free, but certainly less expensive. So a lot of these big shops actually are already dual hypervisor entities anyway. So we've always had to do that. If they decide to, if you will, get sucked into Dell, which I'm not sure would be the right thing, but if that's the decision they make, we want to make sure that our customers are taken care of, and I do think a lot of customers would flee to another hypervisor slash container. If they stay independent and do what they did under EMC's range, and where are we working with them all the time anyway? We work with them in the channel. We have a channel program going on with them at one of our large business partner series. That's gone very well for us. It's a virtualization with our flash systems through our partner. So we'd love to keep doing that. I think it depends. The two and three would be the best, right, Stu, for the industry. I mean, the Switzerland, you know, and or spinning the whole thing out, would be best for the industry, would you agree? Yeah, absolutely. I mean, Eric, you've worked, you know, we spent over the last decade trying to fix all the things that virtualization broke in the storage stack. I'm curious, you've mentioned containers a couple of times. Can you kind of unpack for us where we are kind of in storage for containers? Because most of the early applications were really kind of stateless there. And just kind of where storage fits as to where does it just work, and where does it have to kind of tie in to make, you know, storage work better for containers? Right, well, one of the things from a container perspective, and of course, you had it in the early virtualization, was how do you keep data consistencies and move things around? You know, now you've got vMotion, you know, obviously the storage vendors have all worked with VMware to make sure that that ended up being right, and that data didn't get lost, data was not being corrupted, data was still available. The container guys are going to go through the exact same thing. It may be a different paradigm. I would actually argue that containers are sort of an evolutionary paradigm. You started with the virtual machine, now a container is the next wave, but it's not a radical transformation because the container, again, is the app is sitting in the container and the data is sitting in the container, which is what a virtual machine does. How they, there are definite differences and so from a storage perspective, you want to make sure that, A, if data gets moved from container to container, it doesn't get corrupted, it's still valid, still good. Do you have data consistency? If you're replicating from, you know, if you really, if containers become like VMware and Hyper-V and they, if you will displace virtualization, then they're going to be moving things around, you got to replicate from one site to another site. All those data consistency issues that you had to deal with in the virtualization layer, first with VMware, then with Hyper-V, then with all the other, you're going to probably end up with the same sort of issues in the container space. So we're already working with some container vendors. We've already been engaged with them and make sure that as they get out there and it becomes a bigger business, that, you know, we can handle the storage issues such as data consistency, avoiding corruption on data, move from container to container. The same issues that, quite honestly, IBM had to deal with all the virtualization vendors as well. So a lot of those challenges have to happen in the open source world. You know, IBM's got a great position in open source. Does that give IBM an advantage to go in and be important in kind of the adoption of containers? Well, you know, we are the biggest contributor to the Linux world of any of the existing vendors. And in this case, I'm not talking just storage, I'm talking IBM, the corporation. So we've got more guys working on Linux. You know, we've got a mainframe initiative. We have a power initiative. Obviously, we support OpenStack and all the other frameworks that are out there from a storage perspective as well. So, you know, you do need to be in that space and we're in that space in a big way as a company. So as things need to happen in the container market space, we're there, we have credibility with the developer community, right? And if you will, street cred that the developers say, okay, IBM, well, a commercial entity understands how the open source community works and is fair, equitable, you know, and doesn't try to strong arm them. So I think that that's to our advantage as containers displace, and that's if they do displace virtual, I think it's still an open switch. I don't think there's a real decision yet. You know, when virtualization came out of kind of, it really did sweep the market quickly. Containers have been around for a while now and they don't appear to be knocking all the virtualization vendors on their heels yet. So it'll be interesting to see, can they take their game up a notch and knock all the virtualization vendors out? Do the virtualization vendors get smart? Because if I was running either the Hyper-V division or if I was Pat, I would be saying, I'm going to create my own container thing or I'm going to buy a container company. So that way, if the guy wants containers, I've got the best containerized solution. I've got the best virtualization solution, but I'm not in that business. For us, whatever they choose, any one of the virtual engines, any one of the container engines, we didn't make sure that the data is always there. It's always available. It is not being corrupted and it can be adequately replicated and be consistent when it gets to that other side. Those are the same issues we dealt with virtualization. So yes, we need to be involved because of the openness of the community, but IBM's well-positioned already has been doing that in a number of different areas for the company anyway. I mean, I was, I'm sort of surprised that if something hasn't happened there, maybe the price is too, too rich for Docker or even core OS and they just couldn't justify it. I want to talk about product. One of the things that I was excited about when you came to IBM is you're a product guy and you understand that storage has always been about cycles, getting product to market, getting R&D into the product pipeline, into customers' hands. So give us the rundown on the product portfolio. What's new? What's hot? So this year for us has been an incredible amount of product launches. So we started with early flash launches. We introduced the A9000 family, store-wise all flash arrays. We've got the deep flash 150 with spectrum scale for big data and analytics. So a huge raft of products around that. We also own the software-defined computing part, what we call spectrum computing. And so we launched a distribution that included PatchySpark, our spectrum computing software and our spectrum scale, again for big data analytic workloads. So we had a lot of take up there. We also refreshed that spectrum computing line in July, comprehensive first sort of major overhaul and in some cases as much as 80 times the throughput from what we were doing, because we hadn't done a major overhaul for three years. So an incredible boost in performance, flexibly in fact, earlier today at Tom Rose Emilia's Edge leadoff session, one of the customers for spectrum computing is Red Bull Racing. And Red Bull Racing was talking about how they use it and how they use it to control all of their work flows, et cetera. And then the last thing of course has been the spectrum storage family. We launched a suite. So up to 40% discount that was in January. We had a major launch with spectrum scale, which now transparently teals tiers to any cloud with the REST API, to IBM Cloud Object Storage, formerly known as Cleversafe, Spectrum Protect, Backsup Cloud. So we actually did a launch last week. We already had IBM Object Storage, Cleversafe support for backup as source or target. We could see SoftLayer as a target. We added Amazon S3 last week. Okay, so we're continuing from a cloud perspective and we have close to 400 smaller CSPs and MSPs that already use Spectrum Protect for backup. So a whole bunch around that and we just announced copy data management today. So we now have a copy data management solution. As you know, you can generate all these copies, whether it be the snapshot, the replicas, how do you manage it, how do you catalog it, how to essentially, how do you keep it from getting out of control. And when you get out of control, originally as you remember, it was all about replication for disaster recovery only in case it was an earthquake, fire, or flood. Now you're replicating to give a real live data set to the developers to work on. So now you've got copies galore and then you have snapshotting of course, which allows you to help your backup, the snapshot is the source that's set in the primary server. So the copy data manager is going to be a very rich market for us and we just launched that product today, Spectrum copy data management. So that will be available in early October. So really fleshed out the Spectrum line. We special a lot around all flash, the Spectrum computing line, and then lastly our VersaStack takes advantage of what we've done with all of the flash array technology from both the flash system side, which uses custom flash modules, what we've done with our store wise family, we have all flash across our VersaStack line as well, which we do jointly with Cisco in the conversion infrastructure space. So there's a lot and there's more coming. We have a major launch coming of our Spectrum virtualize technology coming in the October timeframe. So that would be very cloud centric. So hopefully people will tune in. So we've got stuff coming there. We have some additional things happening in flash also in Q4. So very, very busy. And the key thing is people don't buy products, storage guys like me, I love that stuff. But you know what? Every CIO, every CFO, even half the guys in IT, they hate storage. It's about applications workloads and use cases. You ought to show how your storage makes their applications workloads and use cases better because they're all about Spark or if they're traditional, they're all about Oracle or SAP. And you really don't care. I have yet in 31 years of storage to ever met a CIO who was a storage admin ever. And I've done thousands of executive briefings between IBM the first time, between EMC, all the startups, run around with channel partners all the time. I have never, ever, ever met a CIO that was a storage guy ever. So they're mostly software guys now. And so they don't like to deal with storage. And it is about applications workloads. It's all about the data. How do you use the data? How do you take care of that data? How do the apps use the data? So the storage vendor that is making sure they work in the healthcare space and got right things for the financial application base, right things for government applications. That's what matters. And you know, the storage guy who understands how to solutionize their storage products and not just be storage speeds and feeds. That's the future of where things, and you know, it's a very tight market for all the storage vendors. I mean, it's been a tight market the last three years. Flash has really impacted the revenue base for vendors like us. The more flash you sell, you don't sell the big $5 million boxes short-stroking hard drives. You sell an all-flash array for $300,000 and software defined. And we were again, one of the analyst firms that of course tracks the numbers. Their numbers came out last Thursday and when you take the storage system, business and storage software added together, we're the number two storage company in the world. But our software business is skyrocketing, which actually means we had a lower number in hardware, but the software went up. But we don't sell as much hardware anymore. We sell more storage software and software defined. That was the IDC numbers? Yeah, the IDC numbers came out last Thursday. So their storage systems and their storage software came out when you take storage systems, which is what is for us. The storage company is both. Some guys aren't. I mean, we have some smaller players that are software only. You got some all-flash guys that only do flash arrays. We do both. When you add it together, we are the second biggest storage company in the world again and have held that position now for the last five years. Behind EMC. Well, Dell now is number one. Were they counting EMC and Dell together? Yeah, Dell and EMC were merged together for the first day. Oh, no, they were not, because it was last quarter's data. So right, they're always a quarter behind. So in Q2, they were separated out. So we were, we've been number two now for five years in a row, adding the complete storage solution set. We're not a storage software vendor only. We're not a storage system vendor. We have all those solutions. And of course, if you think about it, a lot of the value out of our arrays, Spectrum Accelerate comes on the FlashSystem A9000 and the XIV. Spectrum Virtualize comes on the FlashSystem V9000 and all the store-wise product lines. So a lot of the value is in those two pieces of software. And then of course, you need backup and archive to protect it. And we did very well in, done very well with Spectrum Protect, which again, according to the IDC guys, is number one in enterprise backup, or number two overall, that you got guys like, you know, they're playing in the low end like Veritas, but for enterprise backup, we are the number two and number one enterprise backup player in the world. And number two overall, cause you don't really have low end backup. We only have enterprise. So a lot of activity going on this year and a lot of launches. So- Love it. And you're right. CIOs that care about outcomes, you know, the infrastructure to them is plumbing. Right. They want it to be rock solid, dirt cheap and super fast. And then don't talk about it. And it never needs to fail. Right. Always available. Oracle Open World's going on this week, head to head with Edge. What are you guys doing in the Oracle world? So IBM actually has a giant booth there. We're a grand sponsor, which is one of the top levels. All the division, not every division, but most divisions of IBM are going, the storage division actually has more pedestals than anybody. So we have our DS 8888, all flash mainframe attached. You know, 2.5 million IOPS, about 500 microseconds of latency. But of course, most importantly for that, those workloads is reliability, availability and serviceability, six, nine box, 31 seconds of downtime a year, multi-site replication, you know, all the things that are important in that space. We have all our software defined products there and obviously flash. Right. Our other flash systems that are not in the mainframe space. Oracle for the storage division is probably our number one or number two workload of all the workloads. They're not counting virtualization because obviously a lot of the world is virtualized, but to me, that's an infrastructure too. Virtualization is not an application. Oracle, SAP, you know, frameworks like Spark, databases like Hadoop or Mongo, right, those are the apps. So from an app perspective, Oracle, if not our number one is definitely our number two application workload. So we've got a nice presence there and we've got a lot of our end users there. Some end user here obviously at this show, but we have a bunch of end users. We actually have some that are both. Part of the guys are here and part of the guys are over there. So we're actually having meetings in both locations at the booth at Oracle World and then also here at Edge. Great. Industry's changing quite dramatically. Dell EMC, you got Amazon coming in. China's another one. You have a background in Chinese history. So, and nobody's really talking about China. You know, it was interesting to see how China held up that Dell EMC deal. It'd be interesting to find out what concessions you know, were extracted from Michael for the China deal. But clearly China is increasingly a global power. How do you see, again, with your background in Chinese history, what do you see as this pontificates Ramon? China's role in this global economy from an IT perspective. So from an IT perspective, obviously right now their focus has been more at the low and mid range. Okay, so desktops, what's left of the desktop world, the laptop world, obviously a lot of the big vendors use the, you know, ODM original design manufacturers and then they actually bring it out with their logo on it. So they've already been very powerful there. They've started to add storage. Lenovo's a partner of ours and OEMs and it leverages our product in the field. We have a partnership with another Chinese vendor who's going to be using some of our software technology. So for us we see it as a good market. The guys who do track the number space, we've been the number one all flash ray in China for the last two years. Obviously this year's not over, but for 15 and 14, we were the number one all flash ray in China. So that's been good for us. I do think what are the Chinese really going to do in storage? Are they going to buy somebody? Are they going to try to do it on their own? Sometimes they are overly sensitive to the cost factor. Not that cost isn't important, but let's take flash. Is flash basically the same price as a hard drive array now? Yeah. Was it more expensive two years ago? Uh-huh. But because you could reduce your server farm because you didn't have to short stroke 1,000 hard drive based arrays, the impact on the power, the rack, the stack, the floor, the servers and in the mid-size where you don't always have all the software enterprise license agreements, software licensing for mid-size companies. So even when flash was quote more expensive cost per gig for the data center was way cheaper. So, but do the Chinese understand the impact of flash and are they waiting? And you know, also what about the resiliency characteristics? You know, the products that have come out of China so far that have been native done. The resiliency characteristics, the five nines is six nines have been kind of mediocre. What about replication? I mean, you really need those technologies to have a real robust, whether it be a traditional data center, obviously, in today's modern world, the cloud. So they could be a factor. They could buy somebody. Are they going to understand the nuances that it's not just about low cost. It's about cost effective and delivering value for that cost. And sometimes they tend to be, it's just, if you excuse my French, cheap. And cheap is maybe it's okay at Herzog's Barn Grill for global Fortune 2000 enterprises. Cheap is usually not good. Cost effective, delivering the right performance levels, giving the right reliability, availability, and serviceability is critical in these top 2000 companies in the world. It's critical. And right now they don't appear to have understood that well enough to be a force. That said, the Chinese always end up being a force long term. So IBM has said, hey, we're going to swim upstream. Software defined, flash, you know, cloud, cognitive, analytics. That's the tailwind that we're going to ride. Higher margin, value, business value, business outcomes. You've sort of made that bet. Good bet. It's worked for IBM in the past. It's likely work in the future. The industry though is changing. You know, the days of 60% gross margins, you know, as an independent storage company for EMC and NetApp seem to be shifting. I don't know, maybe Peor can replicate that. Maybe others, but. I think the key thing is that's part of the reason why a margins will be tight in the system side. They're not going to be what they were in the past. The cloud, whether it be public cloud in the low end or hybrid cloud for the bulk, you know, if you, I'd say upper-mid market, you know, the $5 billion company app, most of those guys are not going to go 100% cloud. They're going to go hybrid. Now by the way, that couldn't mean half their workloads are sitting in clouds. Okay, so you're only going to have half of the infrastructure that you need on site. Same thing, flash has dramatically lowered the price. You don't need a $5 million array, short stroke it, try to get 200,000 IOPS and boy, four milliseconds of latency, you think you died and go to heaven. Our all-flash array for 300,000 is a million on a single array, and you can aggregate the arrays, but on a single array, a million and a hundred microseconds of latency, which, and of course, latency is the number one enemy of most applications. So that right away has just ripped the pricing out of the storage business. Cloud is a factor, and one of the good things about software defined, and you know, what we do is to steal from Burger King, have it your way. You want it on-premise, we have full arrays. You want cloud, we have cloud, we have IBM software, but the IBM storage division works within and a ton of different cloud providers. Like I said, Spectrum Protect is available from close to 400 MSPs and CSPs. We sell our storage tape to other cloud providers. So you've got cloud versions, you have full array versions, and you also have software-only versions. So from our perspective, we're riding that wave, and I think one of the things that IBM has gotten great checkmarks, great kudos for in this pass of this change in transformation is it is have it your way. You want cloud, we got all this stuff in the cloud for storage. You want to still get an on-premise array, great. The A9000, for example, has features that are cloud-centric. Secure multi-tenancy, thresholding, and quality of service. Now, we could sell that to a cloud service provider. We then would have Flash in their data center or to a hybrid cloud for the enterprise guys. The point is, you can a full array, in this case, with cloud feature sets, cloud-only configs, and then software-only, which doesn't have to work on hardware. Spectrum Virtualize, which is a block virtualization stack, which provides enterprise-class data services, SNAP, replication, real-time compression, data at rest encryption, et cetera. We support close to 400 arrays. They don't have an IBM logo. So if they want to buy software-only, God love them. They want to buy cloud, God love them. They want to buy arrays, God love them. You know what? The big enterprises often may buy all three. This division uses this, these workloads are on-prem, this workload is software-only, so we're seeing some of our biggest customers actually consume the storage in all three ways. So for us, we have that flexibility. A lot of other guys don't. Right, and increasingly, do you see a lot of growth in just surfacing those services through APIs, I mean longer term, that whether it's in the cloud or wherever, and you're charging for those services as they're consumed, is that? Right, so in that case, if you will, storage is a service model, and we've got that available today with Spectrum Accelerate, Spectrum Protect, obviously through these cloud providers. So we view it as, it will transform over time, but in the short term, or actually I'd say to the midterm, people want flexibility and we have three different instantiations of consumption and almost nobody else does. They don't have software-only, they don't have cloud versions, someone's a software-only company, some guy's an array, pure's a nice little company, but they're array-only and they're losing money. They have no cloud configuration. They don't have a software-only load that you can just buy the software and buy the flash from anybody else, or buy the flash from us and use their software. We do. You can take Spectrum Accelerate, Spectrum Virtualize, Spectrum Protect, you have a whole set of products that work with everybody else's gear that's not ours, and in many cases, available as cloud configurations, which is exactly, so it gives the customer freedom of choice and a lot of the other vendors don't have that capability or technical depth to be able to do that, and we can. All right, we got to leave it there, but I'll give you the last word, put a bumper sticker on what you just said. Why should I buy from you? We've got choice and we are writing the wave of the future, not looking at the past. Shoulda worn my Hawaiian shirt. Ha ha ha, Eric Herzog, what always a pleasure, man, seeing you. Thank you. Thanks very much for coming up. Thanks too. Keep right there, everybody. Stu and I will be back. This is IBM Edge. We're live from Las Vegas, be right back.