 Live from Barcelona, Spain, it's theCUBE. Covering Cisco Live 2018. Brought to you by Cisco, Veeam, and theCUBE's ecosystem partners. Hello everyone, welcome back. This is theCUBE Live here in Barcelona for Cisco Live. You're up, I'm John Furrier, the co-host of theCUBE with Stu Miniman, analyst at Wikibon covering networking, storage, and all infrastructure cloud. Stu Miniman, Stu, our next guest is Eric Herzog, who's the chief marketing officer at IBM Storage Systems. Eric, CUBE alumni, he's been on so many times, you can't even count, you get the special VIP badge. We're here bringing down all the top stories at Cisco Live in Europe. Kicking off 2018, although it's the European show, not the big show, certainly kicking off the year with a lot of new concepts that aren't necessarily new, but they're innovative. Eric, welcome to theCUBE again. Well, thank you. We always love participating in theCUBE. IBM is a strong supporter of theCUBE and all the things you do for us, so thank you very much for having us again. A lot of great thought leadership from IBM, really appreciate you guys' support over the years. But now we're in a sea change. IBM had their first quarter of great results, and that we well report on SiliconANGLE, but the sea change is happening. You've been living this generation, you've seen a couple of cycles in the past. Cisco putting forth a vision of the future, which is pretty right on. They were right on at Internet of Things 10 years ago. They had it all right, but they're a networking company that's transformed up the stack over the years. Now on the front lines of no perimeter, more security challenges, cloud, big whales at no networking and storage, you're in the middle of it. Break it down. Why is Cisco Live so important now than ever before? Well, for us, it's very important because one, we have a strategic relationship with Cisco. The storage division does a product with Cisco called the VersaStack, converged infrastructure, and in fact, one of our key constituents for the VersaStack are MSPs and CSPs, which is a key constituent of Cisco, especially with their emphasis on the cloud. Second thing for us is IBM Storage has gone heavily cloud. So going heavily cloud with our software, in addition to what we do with our solutions as the foundation for CSPs and MSPs, just what we've integrated into our software to find storage for cloud, makes Cisco Live an ideal venue for us and Cisco an ideal partner. So I got to ask you, you've had conversations in theCUBE before, they're all on YouTube.com, so just look at the angle, just search Eric Herzog, you'll find them. But I want to recycle this one point and get your comments and reaction here in Barcelona. You guys have transformed with software at IBM, big time with storage, okay? You position well for the cloud. What's the most important thing that companies have to do like IBM and Cisco to play an innovative role in the cloud game as we have software at the center of the value proposition? Well, I think the key thing is when you look at cloud infrastructure, first of all, the cloud's got to run on something. So you need some sort of structural infrastructure foundation, servers, networking, and compute. So at IBM and with Cisco, we're positioning ourselves as the ideal rock solid foundation for the cloud building, if you will. So that's item number one. Item number two, our software in particular can survive not only on-premises, but can bridge and go from on-premise to a public cloud creating a hybrid infrastructure. And that allows us to also run cloud instantiation. Several of our products are available from IBM Cloud Division. Amazon offers some of the IBM storage software. Over 300 cloud service providers, smaller ones offer IBM Spectrum Protect as a backup service. So we've already morphed into storage software, either A, bridging the cloud in a hybrid config, or being used by cloud providers as some of their storage offerings for end users and businesses. Yeah, Eric, I want to dig into one of the partnership areas that you talked about with Cisco is versus stack. You know, we've talked with you a number of times about converged infrastructure, that partnership, Cisco UCS, taking all the virtualization. The buzz in the market, there's a lot of discussion, oh, it's hyper-converged, it's cloud. You know, why is converged infrastructure still relevant today? Well, when you look at the analysts that track the numbers, you can see that the overall converged market is growing and hyper-converged is viewed as a subset. When you look at those numbers this year close to 17 billion U.S., about 75% of it is still standard converged versus hyper-converged. One of the other differences, it's the right tool for the right job. So customers need to go in, eyes open. So when you do a hyper-converged infrastructure, by the way, IBM offers a hyper-converged infrastructure currently with Nutanix. So we actually have both, the Nutanix partnership offering hyper-converged and a partnership with Cisco on standard converge. It's really, how do you size the right tool for the right job? One of the negatives of hyper-converged, very easy to deploy, that's great. One of the negatives is, every time you need more storage, you have to add more server. Every time you need more server, you add more storage. With this traditional converged infrastructure, you can add servers only, or networking only, or storage only. So I think when you're in certain configurations, workloads and applications, hyper-converged is the right solution, IBM's got a solution. In other situations, particularly as you're middle-sized and bigger apps, regular converge is better because you can basically parse and size up or down, compute, networking, and the storage independent of each other. Whereas in hyper-converged, you have to do it at the same time. And that's a negative, and where you're either over-buying your storage when you don't need it, you're over-buying your compute when you don't need it. With standard converge, you don't have that issue. You buy what you need when you need it. But I think most big companies for sure have certain workloads that are best with hyper-converged, and we've got that. And other workloads that are best with converged, and we have that as well. Okay, the other big growth area in storage for the last bunch of years has been Flash. IBM's got a strong position in all Flash arrays. What's new there? How are some of the technologies changing? Any impact on the network that we should be really understanding at this show? Sure, so a couple things. So first of all, we just brought out some very high density all Flash arrays in Q4. We can put 220 terabytes in two RACU, which is a building block that we use in several different of our all Flash configurations, including our all Flash versus staff. The other thing we do is we embed software-defined storage on our, software-defined storage, actually on our physical all Flash arrays. Most companies don't do that. So they've got an all Flash offering, and if they have a software-defined offering, it's actually a different piece of software. For us, it's the same. So it's easier to deploy. It's easier to train. It's easier to license. It's easier for reseller to sell if you're happy to be using a reseller. And the other thing is it's battle-hardened because it's not only standalone software, but it's actually on the arrays as well. So from a test infrastructure quality issue versus other vendors that have certain software that goes on their all Flash array, and then a different set of software for all software-defined. It doesn't make logical sense when you can cover it with one thing. So that's an important difference for us and a big innovator. I think the last thing you're going to see that does impact networking is the rise of NVMe over fabrics. IBM did a statement of direction last May outlining what we're doing. We did a public demonstration of an Infiniband fabric at the AI summit in New York in December, and we will be having an announcement around NVMe fabrics on the 20th of February. So stay tuned to hear us then. We'll be launching some more NVMe with fabric infrastructure at that time. Yeah, Erica, just for, you know, people that have been watching, you know, there's been a lot of discussion about NVMe for a number of years, NVMe over fabric, you know, more recently, how big a deal is this for the industry? You've seen many of these waves, you know, is this transformational or is it, you know, every storage company I talk to is working on this, so how is it going to be differentiated? What should users be looking to be able to, who do they partner with? How do they choose that solution and when's it going to be ready? So first of all, I view it as an evolution. Yeah. Okay, if you take storage in general, arrays, you know, we used to do punch cards. I'm old enough, I remember using punch cards at the University of California. Then it all went to tape. And if you look at old Schwarzenegger movies from the 80, I love Schwarzenegger spy movies. What's there? IBM systems with big IBM tape and not for backup for primary storage. Then in the late 80s, early 90s, IBM and a few other vendors came out with hard drive based arrays that got hooked up to mainframes and then obviously into minis and to the rise of the land. Those have given away to all flash arrays. From an connectivity perspective, you had scuzzy. You had ultra scuzzy. You had ultra fast scuzzy, ultra fast wide scuzzy. Then you had fiber channel. So now as an infrastructure both in an array as a connectivity between storage and the CPUs used in an array system will be NVMe. And then you're going to have NVMe running over fabrics. So I view this as an evolution, right? What's the driver? Performance or flexibility? A little bit of both. So from the inbox perspective, inside of an array solution, the major chip manufacturers are putting NVMe to increase the speed from storage going into the CPUs. So that will benefit the performance to the end user for applications, workloads and use cases. Then what they've done is Intel has pushed with all the industry, IBM's a member of the NVMe consortium as well has pushed using the NVMe protocol over fabrics which also gives some added performance over fabric networks as well. So you've got it. But again, I view this again as evolution because punch cards, tape was faster, hard driver arrays were faster than tape, then flash arrays are faster. Now you're going to have NVMe in the flash array and also NVMe over fabric for without connecting all flash arrays. So I have to ask you the real question that's on everyone's mind that's out there because storage is one of those areas that you never see it stopping. There's always venture backed startups. You see new hot startups come out of the woodwork and there's been some failures lately in some blame NVMe's innovation to kind of killing some startups. I won't name names, but the real issue is the lines that were once blurred are now forming and there's the wrong side of history and the right side of history. So I got to ask you, what's going to be the right side of history in the storage architecture that people need to get onto to win in the future? So there's a couple of key points. One, all storage infrastructure and storage software needs to interface with cloud infrastructure. Got to be hybrid. If you have a software play like we do where the software such as our Spectrum Scale or our Spectrum Protect or Spectrum Connect Plus can exist as a cloud service through a service writer, that's where you want to be. You don't want to have just a standard array and that's all you sell. So you want to have an array business. You want to make sure that's highly performant. You want to make sure that's positioned in the infrastructure underneath clouds, which means not only very fast, but also incredibly resilient. And that includes both cloud configs and AI. If you're going to do real-time AI, if you're going to do dark trading on Wall Street using AI instead of human beings, A, if the storage isn't really fast, you're going to miss a $10 million, $100 million transaction. Second thing, if it's not resilient and always available, you're really in trouble. And God forbid when they bring AI to healthcare, and I mean AI in the operating room, boy, if that storage fails when I'm on the table, wow. That's not going to be good. So those are the things you've got to integrate with in the future, AI and cloud, whether it's software-defined in the array space or if you're like IBM in both markets. Performance and resilience. Performance and resiliency is critical. All right, so Eric, I have a non-storage question for you. Absolutely. So you've got the CMO hat for division of IBM. You've been CMO of a startup, been in this industry for a while. What's the changing role of the CMO in today's digital world? So I think the key thing is digital is a critical method of the overall marketing mix. And everything needs to reinforce everything. So let's take an example. One of the large storage websites and magazines recently announced that IBM is a finalist for four product of the year awards. Two for all flash arrays and two for software-defined storage. So guess what we've done? We've amplified it over LinkedIn, over IBM Facebook, through our Twitter handle. We leverage that. We use it at trade shows. So digital is A, the first for A, right? People look on your website and look at what you're doing socially before they even decide, should I really call them up or should I really go to their booth at the trade show? So discovery and learning is happening online. Discovery and learning, but even progression. We just, I just happened to tweet and LinkedIn this morning. Clarinet, a large European cloud MSP and CSP just selected IBM all flash arrays, IBM Spectrum Protect and IBM Spectrum Virtualize for their cloud infrastructure. And obviously their target, they sell to end users and companies, right? But the key thing is we tweeted it. We linked it in. We're going to use it here at the show. We're going to use it in PR efforts. So digital is a critical element of the marketing mix. It's not a fad. It also can be a lead dog. So if you're going to a trade show, you should tweet about it and link it in. Just the way you guys do. We all knew you were coming to the show. We know you're going to IBM Think. We know you're going to VMworld and Oracle, all these great shows. How do we find out? We follow you on social media and on the digital market space. And video, video big role. Video is critical. We use your videos all the time, obviously. I always tweet them and link them in once I'm posted. Clip and stick is the new buzzword. Clip them and stick them. Our new Clipper tool. You've seen that. Yes, I have. So it's really critical though that, you know, you can't, and remember, I'm like one of the oldest guys in storage business. I'm six years old. I've been doing this 32 years, seven startups, EMC, IBM twice, Maxdoor Seagate. So I've done big and small. This is a sea change transformation in marketing. The key thing is you have to make it, not stand on its own. Integrate everything. PR, analyst relations, digital in everything you do, digital with shows, and how you integrate the whole buyer's journey and put it together. And people are using digital more and more fast. In fact, I saw a survey from a biz school. 75% of people are looking at you digitally before they ever even call you up or call one of your resellers if you use the channel to talk about your products. That's a sea change. You guys do a great job with content marketing. Hats off to you guys. All right, final question for you. Take a minute to just quickly explain the relationship that IBM has with Cisco and the importance of it, specifically what you guys are doing with them, how you guys go on a market to customers, and what's the impact to the customer? So first of all, we have a very broad relationship with Cisco. Obviously, I'm the CMO of the storage division, so I focus on storage, but several other divisions of IBM have powerful relationships. The IoT group, the collaboration group. Cisco's one of our valued partners. We don't have networking products, so our global technology services division is one of the largest resellers of Cisco in the world, whether it be networking servers, converge, what have you. So it's a strong, powerful relationship. From an end user perspective, the importance is they know that the two companies are working together hand in glove. Sometimes you have two companies where you buy solutions from A and B, and A and B don't even talk to each other, and yes, they both go to the PlugFest or the Compatibility Lab, but they don't really work together, and their technology doesn't work together. IBM and Cisco have gone well beyond that to make sure that we work closely together in all of the divisions, including the storage division with our Cisco-validated designs. And then lastly, whether it's delivered through the direct sales model or through the valued business partners that IBM and Cisco share, it's critical that the end user know, and the partners know, they're getting something that works together and doesn't just have the works option. It's tightly honed and finally integrated, whether it be storage or the IoT division, the collaboration division. Cisco is a heavy proponent of IBM security division. Product teams work together. Yeah, they all, all the product teams work together, trade APIs back and forth, not just doing the, let's go do a test compatibility test, which everybody does that. But we go well beyond that with IBM and Cisco together. And it's a key relationship for you guys. Key relationship for the storage division as well as for all the, many of the other divisions of IBM, it's a critical relationship with Cisco. Eric Herzog, Chief Marketing Officer for the Storage Systems Group at IBM. It's theCUBE, live coverage in Barcelona. I'm John Furrier, Stu Miniman, back with more from Barcelona. Cisco live, you're up after this short break.