 Live from Las Vegas, it's theCUBE. Covering InterConnect 2017, brought to you by IBM. Okay, welcome back everyone live here in Las Vegas. This is theCUBE's coverage of IBM's InterConnect 2017. I'm John Furrier, my co-host Dave Vellante, our next guest is Eric Herzog, CUBE alumni, vice president of product market at IBM Storage. Welcome back to theCUBE. Good to see you with the shirt on. Get the IBM tag there. Look at that. Well, you know, I've worn a Hawaiian shirt now. I think 10 cubes in a row. So I got to keep the streak going. So pretty sunny here in Vegas. Great weather. Storage is looking up as well. Give us the update. Obviously, this is never going away. We talk about all the time, but now cloud more than ever. A lot of action happening with storage. Storage of the data is a big part of it. Yeah. The big thing with us has been around hybrid cloud. It's our software portfolio, the spectrum family, spectrum virtualized, spectrum protect our backup package, spectrum scale our scale out. And as IBM cloud object storage, all will move data transparently from on premises configurations out to multiple cloud vendors, including IBM Bluemix, but also other vendors as well. That software is embedded on our array products, including our versus stack. And just two weeks ago at Cisco live in Melbourne, Australia, we did a benefit amounts with Cisco around our versus stack for the hybrid cloud. So what's the hybrid cloud equation look like for you guys right now? Because it is the hottest topic. It's almost like brute force everywhere as you see is hybrid cloud. That's what people want. How does it change the storage configurations? What's the solutions look like? What's different now than it was a year ago? I think the key thing you've got to be able to do is make sure that the data can move transparently from an on premise location or private cloud. You could have started as a private cloud config and then decide it's okay to use a public cloud with right security protocols. So whether you've got a private cloud moving to a public cloud provider like Bluemix or an on premises configuration moving to a public cloud provider like Bluemix, the idea is they can move that data back and forth. Now, with our Cisco announcements, Cisco with their cloud center is also providing the capability of moving applications back and forth. We move the data layer back and forth with spectrum virtualized or IBM's copy data management products, spectrum copy data management. And with cloud center or the ECS, the enterprise cloud suite from Cisco, you can move the data back and forth, sorry, move the application layer back and forth with that configuration on our versus tax. So this whole software defined thing starts, started when people realized, hey, we can run our data centers kind of the way the big hyperscalers do. IBM pivoted hard towards software defined. What's been the impact that you've seen with customers? Are they actually, I mean, there was a big branding announcement with spectrum everything while back. What's been the business impact of that shift? Well, for us, it's been very strong. So if you look at the last couple of quarters, according to the analysts that track the numbers, from a total storage perspective, we've moved in the number two position and have been now for the last two years. And for software defined storage, we're the number one provider of software defined storage in the world and have been for the last three years in a row. So we've been continuing to grow that business on the software defined side. We've got scale up block configurations, scale out block configurations, object storage with IBM cloud object storage and scale out NAS and file with our IBM spectrum scale. So if your file, block or object, we got you covered and you can use either AR competitors storage, we work with all our competitors gear, or you could go with your reseller and have them or your distributor provide the raw infrastructure, the servers, the storage, flash or hard drives and use our software on top to create essentially your own arrays. So when you say competitors gear, you're talking about what used to be known as the sand volume controller and now is spectrum virtualized, right? Yes, well we still sell the sand volume controller. When you buy spectrum virtualized, it comes as just a piece of software. When you buy the sand volume controller as well as our flash system V9000 and our store wise V7 and V5000s, they come with spectrum virtualized preloaded on the array. So we have three ways where the arrays preloaded, sand volume controller, flash systems V9000 and then the store wise products. So it's preloaded or you can buy the standalone software spectrum virtualized and put it on any hardware you want either way. I know where IBM conference, IBM hates, they don't talk about the competition directly but I have to ask the competitive question. So you've had a lot of changes in the business. Obviously cloud's coming in a big way. The Dell EMC merger has dislocated things and you still see a zillion startups in storage, which is amazing to me, right? Everybody say, oh storage is dead but then all this VC money still funneling in and all this innovation. What's happening in the storage landscape from your perspective? Well, I think there's a couple of things. So first of all, software defined has got its legs now. When you look at it from a market perspective, last quarter ended up at almost 400 million which put on, yeah, let's say a $1.6 to $2 billion trajectory for calendar 2017 out of a total software market of around 16 billion. So it's gone from nothing to roughly 2 billion out of 16 billion for all storage software of all various types. So that's hot. All flash arrays are still hot. You're looking at right now, last year, all flash arrays end up at roughly 25% of all arrays shipped. They're now in price parity. So an all flash array is not more expensive. So you see a lot of innovation around that. You're still seeing innovation around backup, right? You got guys trying to challenge us with our Spectrum Protect with some of these other vendors trying to challenge us, even though backup is the most mature of the storage software spaces, there's people trying to challenge that. So I'd say storage is still a white hot space. Because you know, the overall market is flat. So it is totally a drag out, knock down fight. The MMA and the UFC guys got nothing on what goes on in the storage business. So make sure you wear your flat jacket if you're a storage guy. Meaning you got to gain share to grow, right? Yes, and it's all about fighting it out. This Hawaiian shirt looks Hawaiian, but just so you know, this is Kevlar. Just in case there's another storage coming here at the show. So what are the top conversations now with storage buyers? Because we saw Kenny's announcement about the storage object store flex for the cold storage, changes the price points. It's always going to be a price sensitive market, but they're still buying storage. What are those conversations that you're having? You mentioned moving data around. Do they want to move the data around? They want to keep it at the edge? Is it moving the application around? What are some of those key conversations that you're involved in? So we've done a couple of innovative things. One of the things we've done is worked with our sales team to create what we call the conversations. You know, I've been doing this storage gig now for 31 years, seven startups, IBM twice, EMC, Mac Store and Seagate. You're a hardened veteran. I am a storage veteran. That's why this is a Kevlar Hawaiian shirt. But no CIO is a storage guy. I've never met one. In 31 years, ever, ever, ever met a storage guy. So what you have to do is elevate the conversation when you're talking to the customer about why it's important for their cloud, why it's important for machine learning, for cognitive, for artificial intelligence. You know, think about it, I'm a Star Trek guy. I like Star Wars true. But in Star Trek, bones, of course, wands the body. So guess what that is? That's the edge device going through the cloud to a big giant server farm. If that storage is not super resilient, the guy on the table might not make it. And if the storage isn't super fast, the guy on the table might not make it. And while Watson isn't there yet, Watson Health, they're getting there. So in 10 years from now, I expect when I go to the doctor, he's just like in Star Trek, waving the wand. And boy, you better make sure the storage that that wand is talking to better be highly resilient and high performing. How do you define resilient in your terms? So resilient means you really can't have more than 30 seconds, 50 seconds a year of downtime. Because whoever's on the table when that thing goes down has got a real problem. So you've got to be up all the time. And if you take it out of the healthcare space and look at other applications, you know, whether you look at trading applications, data is the new gold. Data is the new diamonds, it's about data. Yes, I'd love to have a mound of gold, but you know what, if you have the right amount of data, it's worth way more than a mound of gold is. You know, you're right about the CIO and storage. They don't want to worry about storage. They don't want to spend a lot of time thinking about it. CIO once said to me, I care about storage like this. I want it to be dirt cheap, lightning fast and rock solid. Now the industry has done a decent job with rock solid, I would say, but up until flash, not really that great with lightning fast and really not that great with dirt cheap. Prices come down for the hardware, but the management has been so expensive. So is the industry attacking that problem, and what's IBM doing? Yeah, so the big thing is all about automation. So when I talked about moving to the hybrid cloud, I talked about transparent migration, transparent movement. That's an automation play. So you want to automate as much as you can, and we've got some things that we're not willing to disclose yet that'll make our storage even more automated, whether it be from a predictive analytics perspective, self-healing storage that actually will heal itself, you know, go out and grab a code load and put the new code on because it knows there's a bug in the old code and do that transparently so the user doesn't have to do anything. It's all about automating data movement and data activity. And so we've already been doing that with the Spectrum family, and that Spectrum family ships on our storage systems and on our VersaStack, but automation is the critical key in storage. So I wonder, does that bring up new, Eric, new KPIs? Like I presume you guys dog food your own storage internally in your own IT. Yes. Are you seeing, cause it used to be, okay, the light screen and the disk drive and this is our uptime or downtime, planned downtime, and just sort of standard metrics that we've known for 30, 40 years. With automation, are we seeing a new set of metrics in KPIs emerge, self-healing, percentage of problems that corrected themselves? Well, you're also seeing things like time spent. So if you look at, go back to the downturn of seven, eight and nine, IT was devastated, right? And as you know, you've seen a lot of surveys that IT spend is basically back up to a 0.8, okay, the pre-08 crash. When you open up that envelope, they're not hiring storage guys anymore and usually not infrastructure guys. They're hiring guys to do DevOps and test dev and do cloud-based applications, which means there's not a lot of guys to run the storage. So one of the metrics we're seeing is how much guys do I have managing my storage or my infrastructure? I used to have 50, now I'm a big bank. Can I do it with 25? Can I do it with 20? Can I do it with 15? And then how much time do they spend between the networking and the storage, the facilities themselves, right? These data center guys have to manage all that. So there are new metrics about what is the workload that my actual human beings are doing, how much of that is storage versus something else? And there's way less guys doing storage as a full-time job anyway because what happened in the downturn? And so automation is critical to a guy running a data center whether he's a cloud guy, whether he's a small shop, and clearly in the Fortune Global 2500, those guys where they've got in-house IT, they've cut back on the infrastructure team and the storage team, so it's all about automation. So part of the KPIs are not just about the storage itself, such as uptime, cost per gig, cost per transaction, the bandwidth, those sorts of KPIs, but it's all about how much time do I really spend managing the storage? So if I've only got five guys now and I used to have 15 guys, those five guys are managing usually three to four to five times more storage than they did in 2008 and 2009. So now you got to do it with five guys instead of 15, so there's a KPI right there. So what about cloud? We heard David Kenney today talk about the object store with that funny name, and then he talked about this cloud-tearing thing and I couldn't stay, I had to get ready for the cube. How do you work with those guys? How do you sell a hybrid story together because cloud is eating away at the traditional infrastructure business, but it's all sort of one big happy family, I'm sure. But how do you work with the cloud group to really drive, you know, make the water level higher for IBM? So, all of our products from the Spectrum family, or not all, but almost all our products on the Spectrum family will automatically move data to the cloud, including IBM Bluemix slash SoftLayer, okay? So our on-premises can do it. If you buy our software only and don't buy our storage arrays, right, don't buy a store-visor, don't buy a flash system, you still can automatically move that data to the cloud, including to IBM Cloud Object Storage. Our Spectrum scale product, for example, scale out NAS and file system, which is very highly used in big data analytics and cognitive workloads, automatically, by policy, will tier data to IBM Cloud Object Storage. Spectrum protect can be set up to automatically take data and back it up from on-premises to IBM Cloud Object Storage. So we've automated those processes between our software and our array family and IBM Cloud Object Storage and Bluemix and SoftLayer. And by the way, in all honesty, we also work with other cloud vendors, just like they work with other storage vendors, right? All storage vendors can put data in Bluemix. Well, guess what? We can put data in clouds that are not Bluemix as well. Of course, we prefer Bluemix. We all have IBM employee stock purchase. So of course we want Bluemix first, but the customer, for whatever reason, doesn't see the light and doesn't go to Bluemix and goes with something else and we want to make sure that customer's happy, right? We want to get at least some of the PO and our Spectrum family and our VersaStack family and all our array family can get that part of the PO. We need the versatility to be on any cloud. We can be on any cloud. So my question for you is, the thing that came out of our Big Data Silicon Valley event last week was the Hadoop was a great example and that's kind of been now a small feature of the overall data ecosystem is that batch and real time are coming together. Right. So that conversation you're having that you mentioned earlier is about more real time than there is anything else, more than ever. Well, and real time gets back to my example of bones on Star Trek, wanting you over healthcare. That is real time. You know, he's got a phaser burn, a broken leg of this and that and then we know how to fix the guy. But if you don't get that from the wand then that's not real time analytics. I mean, just speaking of Star Trek, just how much data do you think the enterprise was throwing off? It's from an IOT standpoint. I'm sure that they had about 100 petabytes. All stored on IBM flash systems are raised by the way. Eric, thanks for coming on real quick. Give you next 30 seconds. Just give the folks a quick update on why. The update on why IBM storage is compelling now more than ever. Well, I think the key thing is most people don't realize IBM's the number two storage company world that has been for the last several years. But I think the big thing is our embracing of the hybrid cloud, our capability of automating all these processes. When they've got less guys doing storage and infrastructure in their shop, they need something that's automated that works with the cloud and that's what IBM storage does. All right, Eric Herzog here inside the queue, Vice President of Product Market for IBM Storage. I'm John Furrier. We'll have more live coverage from IBM Interconnect after this short break. Stay with us.