 From Las Vegas, it's theCUBE. Covering VMworld 2018. Brought to you by VMware and its ecosystem partners. Hello, Verona. Welcome back to theCUBE's live coverage here in Las Vegas for VMworld 2018. I'm John Furrier, the host of theCUBE. With Dave Vellante, we're here at Eric Herzog, CUBE alumni, multi-return guest. He's also the CMO and vice president of Global Channel Sales for IBM Storage Systems. Great to see you. Sport the new Hawaiian shirt, especially for this CUBE. Is that a new shirt? That is a new shirt. Looking good, my friend. I couldn't keep wearing the same Hawaiian shirt, guys. Welcome back. Well, thank you. We love coming to theCUBE and always have a great time with you guys. So, storage is hot again this year, Dave. So, storage just can't go away. It's on the keynotes. You're hearing Gale Singer talking about software defined, not new to IBM. You've been talking about this for ages, it seems like on the CUBE. Really now, the relevancy is high because you see a lot of hype. A lot of hype and storage right now. A lot of new cloud-based something with storage, saying software defined. So, we're back to the game in storage. Give us the update. Well, so, storage is a hot space. As you know, at the field level, whether it be from the account side or in the channel, it's become fiercely competitive again. Everyone is fighting for share. We've grown five out of the last six quarters. A couple of other guys have grown. I probably saw some recent fundings in the startup side. So, storage is hot again. You know, it's sort of slow there for a while, right? Things were a little down. We grew, some other people weren't grown. VCs weren't funding storage. Now they're funding again. So, there's excitement back in storage. You've got to love it. What are you guys doing now? I know you had an announcement recently. You were in Madrid, been following you on Facebook, been traveling. You were in, I think Madrid or somewhere. Mexico City. Mexico, yeah, sorry, Mexico City. You're traveling. What was the announcement about? You guys had some news, some pretty hard news. So, we announced an NVME All Flash Array, incredible performance at mid-tier pricing. We can get up to 10 million IOPS in a mid-range price product and 100 microsecond latency, but we went very different. You know, a lot of guys can do good arrays. We think we can do the best, but a lot of people got good products. One of the key things we did was in bed, in the array, way more software than you normally get. Everyone does an array manager. We put in a backup product. We put in secondary data management so that you can reuse snapshots, replicas or backups for test and dev. We put in support for containers and we embedded all that software and including it with the array. So, we're going to an era where it's not just about, you know, trying to slap some hardware to get it put in an array manager. It's now all about how you survive in a data-driven, multi-cloud world and that means you got to put a lot of software on your arrays and deliver that to the end user and to your channel partners for their end users. That's what they're looking for. Kind of beyond the box. Got to go way beyond the box. So, speaking of the channel partners, I mean, we've talked about this in the past a little bit, but, you know, the five, seven years ago, the majority of the volume went through what I call box sellers, right? Yes. And when the cloud came in in a big way, they were like, uh-oh, we have to transform or we're going to be toast. Some of them probably said, hey, I'm going to retire. I don't need to. I got that big house, the big boat, I'm good. But others in the sweet spot said, we got a pivot. Right. What does that pivot look like and how is IBM helping them transform their business? So, there's a couple things. A, many of them are evolving and including MSP and CSP on their line card as well as traditional sales. It's all about enabling the customers to handle a multi-cloud and hybrid cloud environment. The true private cloud is, as you guys have coined, right at Wikibon. So, it's all about doing that. And so, the partners, A, have to do that for their end users. B, they're now offering many of them. Not that they can't resell IBM Cloud or Amazon or Azure, but, you know, part of the issue most partners have with that is account control. Even if they're an IBM partner, it's still IBM Cloud, right? John, you, Stu, all of the guys at theCUBE could just re-up after the first order directly with IBM Cloud or Amazon or Microsoft. So, they've all started to evolve into these cloud service providers. So, we help them in a couple ways. First of all, cloud requires incredible underlying infrastructure that can't fail because cloud is now your business. So, if it goes down, your cloud goes down, right? That's one. Second thing is all of our storage software is heavily cloudified. So, we can tier to the cloud. We can back up to the cloud. We can put object storage on-premise or in the cloud. We can automatically tier data and see a cloud as a target device, whether it be a backup or for tiering. So, we've done that. So, we can offer the right hybrid cloud to the partner. He goes in pitches hybrid cloud to his end user, and then on the back side, he gets the cloud services business, and our software allows him to put the data in his cloud. So, that's the way we're helping the partner base. Okay, so you mentioned that you included sort of made backup inherent to the system. We're going to talk about modern data protection tomorrow. So, let's just focus on the flash system component. Talk about where you guys came from. You made an acquisition to get into that business. You did very, very well. But then, like IBM, you started to bring in your own IP, your own innovation. Where are we sort of in that roadmap, and where are we headed? What can you share with us? Yeah, yeah. So, when you look at it, what we've done is made sure that we have flash across the portfolio. So, we have very low-class flash. Our store-wise 5030F, you know, as low as $18,000 street price for a five-drive-all flash array. And as you guys have written, flash arrays are not more expensive than hard-driver arrays anymore, right? It's the same price. So, we've evolved. We have an entry product line, mid-tier product line. We've brought out NVMe from a fabric perspective. We were shipping NVMe over an Infiniband fabric, although in each market, back in February, when other eyes were claiming we're the first guy with NVMe, we were shipping for revenue in February, and we demoed it at a couple trade shows last year. Oh, wait, let me stop you. So, you were first in NVMe in the marketplace? We shipped for revenue in February of 2018. We shipped for revenue NVMe over fabric. So, that's when we were shipping for revenue, and we started showing it in Q3 and Q4 of last year. Okay, all right, sorry to interrupt. Anyway, so, you got an entry product. We put it into the mid-range. We've added NVMe technology from an all-flash perspective. We have specialty products. Our all-flash elastic storage server is focused on big data analytics, machine learning, and AI. That's a great, I mean, it's huge. We have customers with hundreds and hundreds and hundreds of petabytes for things like credit card fraud detection, genomic research, oil and gas, all kinds of that sort of market. We have all-flash in the mainframe. So, what we've done is, A, we're looking at every application workload in use case. B, one of the things you've seen, not just from us, from others, just the way there used to be hard drive arrays, entry, mid-range, low-end, blah, blah, blah, and a lot of different offerings, you're seeing the same thing. What has happened is Flash started as just up in tier zero high-performance databases, and now it's evolved to any primary storage workload. And primary storage, to me, means tier zero, one and two probably, which means you don't always have the fastest, highest performance thing at super expense. You have to have a mid-range product, an entry product, you need products to hook to mainframe. So, we've put Flash in the entire portfolio at all the right price points and for different applications workloads in use cases. So, how do you look at the workloads segmentation? How do you segment the market? If you look at it by workloads, I'm sure you do. What are the hot spaces? What's growing? What's big and sort of steady? And what's maybe smaller, but growing like gangbusters? Well, what we're seeing is every primary storage workload. So, the database workloads, the machine learning AI workloads, the traditional business workloads, VDI is now all Flash. I mean, everything's gone off Flash. The only thing that isn't going off Flash is backup and archive secondary dataset workloads, and that is staying either with cloud providers, like our IBM Cloud Object Storage, or it's going to what I call cheap and deep, 7,200 RPM hard drive giant boxes. We sell those with our object storage or to tape. And we still do tape, cheap and deep tape and big enterprise, and you'd be surprised how many very large, very large cloud service riders when you go to their archive, you're going to IBM tape. I got to ask you a question. I asked Ed Walts the same question. Has customers see the hype in storage now? A lot of new startups coming out, well-funded. How do you talk to customers about the question that they're asking themselves is, how do I determine who's really got software to find storage? Is there a test? Is there an observation? Is there a benchmark? Is there a way to identify who's actually done the work? Who can deliver it? Obviously you guys are shipping product, but they're trying to squint through the noise. You know the product's pretty deep. What is that filter? So I think there's a couple of filters. First of all, your question, anyone including us, is all right, there's three major data types, file, block and object. Do you have software-defined storage for each one of those segments, yes or no? The answer is no, then they're probably very nichey. Maybe they got a good product just for NAS or just for object, but if they don't have it all, that's a very nichey sort of strategy. Okay, it's item one. Item number two, and it's relevant to us, so is the software-defined storage the same softwares you have on your arrays, yes or no? The answer is yes, which we do. A, it's easier to use, B, big giant enterprises may have some software-defined storage in this division and traditional arrays in this division, but it's the same exact software. And then obviously, when you can use your arrays software as software-defined storage with any vendor's stuff, not just with our own, it's battle-hardened. All the use cases, all the corner cases, all the stuff that could go wrong has already gone wrong. Spectrum Virtualize, for example, is probably close to a million pieces of software and storage subsystems out there. So if there's a bug, we probably found it. If there's a weirdness with the new version of Oracle 27 when it comes out, guess what? There's so many boxes out there and so many pieces of software out there. The odds of us finding that are pretty high. So I think first thing is, do you have file block and object? Second thing is, is it embedded with the arrays or not? We sell at stand-alone, and here's the other thing. Some people call it software-defined storage, but you really have to buy their array. With us, we sell the same software on the arrays and completely stand-alone software. Give me a P.O. and we will sell you Spectrum Virtualize with no hardware from us. And that works on what kind of storage device? Anything? Anybody. And how popular is that? Is that more to give customers comfort or do they actually really want that? Maybe some of the service provider customers want that, but does the typical enterprise want that or do they want an appliance? So I will tell you that the enterprise market is split. So I go see big giant enterprises for IBM all the time and half of them when you say software-defined storage, look at you like your head is spinning backwards like in that science fiction movie, right where head spins around. That's half of them. The other half go, I can't wait to do it. And the reason is, as you know with the big downturn in seven and eight, IT budgets got cut right and left. So now if you look at the IT spend and IT headcount today, it's back up to where it was, if not slightly higher. But when you open up the envelope and peer inside and say, wow, how many storage guys are in that envelope? Oh, wow, way less. The envelope's full, but there aren't a lot of storage guys. The mix is shifting. And there aren't a lot of server guys or networking guys. So for big entities that have enough staff, software-designed storage works, for big entities or even small entities that have a really good channel partner to help them deploy, that's a good thing. But for a lot of other people, including some of the biggest firms in the world, it's like, well, we only have like five storage guys left. I know we used to talk to us years ago at your other company when we used to have 50 storage guys, we only got five now. So software-defined storage, that's hard to do. And we've made it easy, we have all these recipes and red books. You can't just go there overnight. You've got to do the work. No, and we have all kinds of certification. So this storage is certified, these servers, this J-Bod, this J-Boff, so we make it easy. But at the same time, if you only have five storage guys in a big company, used to have 50, how the hell are you going to rack and stack, put it together? Either A, need to use a channel partner, which some will do, some don't. So then they say, just sell me an array. And then, of course, for us, the mantra is, well, guess what, it's the same software you could buy software-defined if you ever want to do that later. So that's a big difference. My question to you, and on the channel side, you mentioned channel partner, that could be a good thing. How has partnering in the channel changed with the cloud? You have ecosystems now, you got a lot of different touchpoints, I think Alan Cohen called it, playing tennis to playing soccer. You got to work with a lot of different people in this world. Seems like it's a soccer-like configuration. You got a lot of things passing around. How has the channel game changed with cloud? Because cloud, technically, Amazon is showing you, swipe your credit card, you get stuff. So the role of the channels is evolving. Your thoughts and vision. So I think a couple things. So first of all, one is we already talked about a number of the better partners have realized I got to be a cloud provider. Second thing that they've realized is when they sell a storage system, even if they're not a cloud provider, they sure as heck better make sure that the storage can automatically, transparently, and easily move data to the cloud. Whether they sell the cloud solution, whether they use IBM cloud or Amazon or Azure, all of us have partner programs every major cloud provider has one. So do they use that? But if the storage supplier can't easily put the data out, can't easily back it up, can't easily archive it to the cloud, then for a partner, that's a pretty tough sell. So you've got to make it easy for them. So A, some of them have evolved to become cloud providers as a side business. Others, you have to make sure that they're going to end up selling the cloud or the end user's going to buy the cloud anyway. So you better make sure your solution, that means the one from IBM, supports all clouds. And we don't just support IBM cloud, we support everybody. I have a follow-up question on that. The old channel model, remember when I was back on the day, Dave and I was talking about my channel days at HP, but it was back in the day in the 90s. Profit was everything, margin, giving, making sure the dealers and the shine on channel parts make money, selling your gear. Okay, check. But that's now moved with cloud, where value add is important. Intellectual property, it seems to be the new table stake. Guy, you got to check the box on profit or fund some IP that you're delivering to the partner. How do you guys have that conversation? What is that IP? What do you guys bring to the table? Yeah, a couple of things that we do in addition to making sure that all of our systems are what I call cloudified, easily work with anybody's cloud, so that allows them to sell hybrid cloud, whether it's their own cloud on the backside, or IBM cloud, or Amazon, or Azure. So you want, we've done that. Second thing we've done is create a number of multi-cloud solutions. So these use our software. We've created these really nice recipe books that are easy to follow. The partners can use them to say, ah, Wikibon, Silicon Angle, The Cube, do you need blah, blah, blah? Guess what, we can give you cloud data resiliency and we can give you data reasoning in the cloud and then they follow what we call the solutions blueprints, come to you guys and say, look, we can do it for you. And then the partner does it. So we give them these recipe books, what we call the solutions blueprints for multi-cloud, that'll allow them to use the software that comes on the 9100, but also can be sold standalone, to create cloudified solutions and that's the IP we deliver to them. Then they do the deployment, they do the installation, they make, of course, margin off the install and the support and service. They wrap services around it. So they wrap services around, but we give them the recipe so that they can easily deploy it, easily the deployment from an end user perspective, but it also makes it easier for the partner instead of saying, I got to learn how to do it, so no. See this recipe, follow this and you'll get the perfect German chocolate cake. Follow this recipe, you're getting the perfect pizza. And it's the same software, just refactoring them all, just the way a baker has pizza dough and donuts and all kinds of stuff is basically five or six components, right? Sugar, water, eggs, milk and baking, and then you make all kinds of different things with that in different configs. That's what we've done with all the software. That's a very key thing. They could use your ingredients into delivering a solution to the customer that they're engaging with. Well, and the good thing is our software portfolio is exceedingly broad. It really comes down to software defined storage, modern data protection, which we're going to talk about tomorrow, control and management plane. Those three components. Now we have a number of things. We have file block and object software defined storage, right? So you can pick what you want. We have AI based management control. Great, you can use that. Modern data protection, secondary data reuse, archiving. All that software, you can refactor it and do it in the cloud. You can do it on premises. You can make it hybridized. So that's where the value is, all that software, they can refactor it and that's what the solutions blueprints do. Eric, thanks for spending the time coming on the team. Great to always get the knowledge. I mean, a rapid fire product, you know, data sheet in your head. You always have a pulse of it. Final question for you. VMworld 2018, the partner pavilion seems to be rocking this year. We've seen years pass with, okay, heart beats there, but this is pumping. What's the deal with storage this year? Tell the folks out there, why is this such a big topic this year at VMworld 2018? So I think there's a couple of issues. First of all, everyone realized if you have a VMware infrastructure or a cloud infrastructure, people are realizing if the foundation, the building isn't very good, then the building falls down. And so many end users, whether it be in the old, what I'll say, the more mature virtualized market or now in the cloud market have tried to do a true private cloud like you guys have talked about, tried to do a hybrid cloud and then the whole thing falls down because the underlying storage infrastructure is mediocre to poor and the next thing you know, the cloud falls over. And you guys in fact, in your true private cloud, you comment that if you don't have the right infrastructure underneath the cloud, the cloud will fail. And I think people have realized that and so many big companies have stumbled and then it's kind of gotten out. So it's going, okay, I better have the right software, better have the right storage. If I don't do that, the cloud's not going to work. And the point we heard yesterday in the keynote and also commentary in theCUBE is one of the benefits of the cloud is it can't go down. It doesn't go down. There's a lot of opportunities to keep it redundant. So don't let it fall down. You've got the right foundation. Yes, exactly. Eric Herzog, bring in theCUBE signal here inside the booth here, the broadcast booth here on the VM Village, two sets of theCUBE. I'm John Furrier, Dave Vellante. Stay with us for more coverage from VMworld after this short break.