 Hi, and welcome to a CUBE Conversation with Wikibon. I'm Peter Burris, the Chief Research Officer of Wikibon, and our goal with these CUBE Conversations is to try to bring you some of the finest minds in the technology industry to talk about some of the most pressing problems facing digital businesses as they transform in an increasingly chaotic world. We're very lucky today to have a couple of great thinkers, both from IBM. Ed Walsh is the General Manager of Storage at IBM, and Eric Herzog is the one's product management for the storage group at IBM. Welcome to the CUBE Conversation today. It's always nice, thank you for having us. So guys, you've been running around Silicon Valley today telling your story. We got a couple of questions. Wikibon likes to talk about the relationship between data and digital business. A lot of people will wonder what digital business is. We say that the difference between digital business and business is how do you use your data assets? Now, that's a stance that's, I think, becoming a little bit more vogue in the marketplace today, but that means that storage has a slightly different role to play when we think about how we protect, secure, sustain those data assets. Do you subscribe to this? Is that how you're looking at it, and is that relevant to the conversations you're having with customers? I haven't heard it that way, but it actually makes a lot of sense, and you can jump in as well, Eric, but I would say however you look at your data, if a digital business is leveraging their data, it makes a lot of sense. We use different, I would say metaphors. One would be your data assets are kind of your oil, he who refines it gets value, so if you get insights from it. So if you're not using that, you're kind of putting yourself at disadvantage. We also see a lot of what I'll say is established companies getting disrupted by, true disruptors using their technology and insights of data to disrupt incumbency. We'll call the Uber of my business, almost like a verb these days. It's disrupting me. They're using technology against me, so the key thing, the best defense is actually using technology, getting insights, and then driving new business. But data alone, you need the right infrastructure, either on-prem or in the cloud, and put the right analytics and insight to it. So I would agree completely, and I would also say, think about it, 80% of data is behind your data, it's not searchable by the web, it's how to leverage your data assets in combination with other things to get true insights. Outside data, different things on AI, and really get true insights and then map them into your business. So I would agree with that, I haven't heard that way, but I would agree with that said, good definition of a digital business. Well, what we're seeing is, for companies that are really leveraging the data, it's their lifeblood. And the issue is, data is not small anymore. It's oceans of data, whether that be things from the internet of things, grabbing things, for example, all the telcos have sensors all over, all of their assets, and they're trying to keep the telco up and going. And it doesn't just have to be a giant telco. Small companies have reams and reams of data, it's an ocean. And if they're not mining that ocean, if they're not swimming through that ocean correctly, the next thing you know, the competitor disrupts them. And that is their power, is the ability to harness these oceans of data and use that data in a way that allows them to get competitive advantage. So people think of storage as just a way to sort of place your data, but storage can be an active part of how you increase the value of that data and gain insights, as Ed was pointing out. Well, I think we totally agree with you, by the way. I think it's an important point. In fact, the observation that we've made is the difference between data as fuel, or the reason why that sometimes falls down, although I understand it, and I think it's a decent enough metaphor, is that, unlike fuel, data can be reused multiple times. And it makes the whole point that you're bringing up, Eric, about the idea that you combine insights from a lot of different places with your data, and storage has to play an active role in that process. But it also says something about the idea of storage as kind of something that you put over there at standalone. I mean, it used to be, we worried about systems integration a lot. Open systems kind of changed that. We just presumed that it was all going to come together. IBM's been around for a long time and has lived in both worlds. What do you think the role of systems integration is going to be as we think about storage, the need to do a better job of protecting and sustaining our data assets, especially given the speed and uncertainty with which the world's changing and the dependence it has on data these days. You want to take that first? Well, let me give you a real time example. One of the things IBM just introduced last week was a very powerful new mainframe. One of the key tenants of that mainframe is the ability to secure data end to end from the day the transaction starts with no impact. So while they are doing transactions, millions and billions of transactions on the server farm, it's encrypted from day one. But it eventually ends up on storage. That storage has to extend that encryption so that when you put the data at rest, while you're analyzing the data, you've got it encrypted. When you're putting it at rest, it's encrypted. When you pull it back, because you run analytics multiple times the data is encrypted, eventually certain data sets that take finance, healthcare does end up on archive. Well, guess what? It still needs to be encrypted. So that's an example of how the complete systems integration from the server through to primary storage through the archive is just one example of how storage plays a critical role in extending everything across this entire matrix of systems integration, not just one point thing, but across an integrated solution. And of course, in this case, it's secure transactions, it's analysis of incredible amounts of insight. And of course, where the IBM Z mainframe is incredible power and speed, yet at the same time keeping that data safe while it's doing all the analytics. So that's a very strong story. But that's just one example of how storage plays a critical role in this complete integration of data with a full systems infrastructure. And maybe I could add to that. So that's a good example of on-prem that also can host in the cloud. But if you think of system integration, your data is critical, you need access to it to actually do the analytic workload, the cognitive workloads on top of it. It can be on-prem or in the cloud or actually split between. So you do need, now you're relying on your cloud infrastructure to give you that enterprise class, not only performance, but availability. But it also matters, but it's no longer you as an individual company put that together. But it doesn't matter. The infrastructure doesn't matter how to get that performance. Also you mentioned security and protection, which is where IBM's cloud comes in. Well it's interesting to us that a lot, it's almost natural to expect that the public cloud companies are going to do deep integration. And they're talking about going all the way down to FPGAs. As long as they are able to handle or provide a set of interfaces that are natural and reasonable from an overall workload standpoint. I would expect that we'd see the same type of thing happen in a lot of different on-premise systems too. So the notion of integration, I think you guys agree, is an important trend where it's appropriate and where it's adding value and it should not be discounted just because it doesn't comply with some set of, some definition of open this, that, or the other thing as it has in the past. I'll agree too. It's an end-to-end system, especially when you look at availability performance, which you're talking about your asset being your data and getting insights. If it's just sitting there, it's not very valuable. In fact, you could say it's actually an exposure, but if you're leveraging it, getting insights and driving your business, it's very valuable, right? So you just need to make sure that infrastructure you have, either hybrid cloud or in the cloud allows you to do that, right? But security is becoming more and more a big issue. So I would agree. Well, race is the next question. So again, as long as we're focused on the data and the answer and not the underlying hardware as the asset, then I think we're in good shape. But it does raise the next question. As we think about converged infrastructure and hyper-converged infrastructure and storage, compute, network, and other elements coming together successfully, what will be the role of storage in the future? I mean, storage is not just that thing that sits over there with the data on it. It is playing a much more active role in encryption, in compression, in deduplication, in how it prepares data to be used by any number of different applications. What, how do you foresee the role of storage evolving over the next few years? Sure, I can jump in. You want to take a shot? Well, yeah, so I think one of the key things you've got to realize is the role of storage is to sort of offload some things from the primary CPU. So for example, if you've got a notions of data, what if we could track all that metadata for you? So when the system or the cloud looked for data, it could search everything, whether that was 20 million lung cancer pictures, whether that be MRI, whether that be the old-style X-ray, go back 20 years. If all that metadata is attached, then the CPU from a server perspective to run the analytics workloads is offloaded and the storage is performing a valuable function of tracking all that metadata so that when the server does its analytics and then has to reiterate several times, for example, Watson, IBM Watson, is a very intuitive element that analyzes, learns, analyzes, learns, analyzes, and it keeps going to get, and it's used as an oncology, Watson's used in financial services, and so if you could offload that metadata now to the storage where it's actually acting almost as if it's a sub-compute element and handling that offloading the CPU, then more time is spent with Watson looking at the financial data, looking at that medical data, and storage can become a very valuable resource in this future world of this intense data analytics, the machine learning, the artificial intelligence that systems are going to provide on-premises through a cloud infrastructure storage. That's just one example. How storage, as an intelligent storage vehicle, is offloading things from the CPU or from the cloud onto the storage and helping it become more productive and the data be more valuable that much faster. I would agree, and I think storage has always been evolving, right? So there's, storage has gravity, has value. If you think of storage as where you store data, it's going to change architecturally. You mentioned hyperconverge, you mentioned converse, you mentioned cloud, we talked about what we can do in the mainframe. It's all about how do you get the right accessibility and performance, but it will change, it will change rather dramatically. Just think of what's going to go on with, we'll say the traditional modernizing traditional workloads, what you do with VMware, and the arrays are getting much more complex. You can also do a software-defined arrays, which allows you to have just more flexibility in deployments, but in the new workloads, where you're looking at high-performance data analytics or doing things that you can actually expand out and leverage to cloud, that becomes much more of a software-only play. It's still storage. The bits and bytes might be on, it's going to be typically on flash, in my opinion, both on-prem or off-prem, but how do you move that data? How do you keep accessibility? How does you cure that data? How do you make sure you have it in the right place where you can actually get the right performance? And that's where storage is always going to evolve. So it doesn't matter if it's in this array, it's in a file system, in what we'll call a big storage array, or it's in the cloud, it's about how do you monitor and manage that through its full lifecycle. So it sounds like you're suggesting, and again, I think we agree, is that storage used to be the place where you put stuff, and it's becoming increasingly where you run data-related services, whether it's services associated with security, or prepping data, or protecting data, or moving data as effectively as possible. Increasingly, the storage resources are becoming the mechanism by which we are handling these strategic data services, is that right? Yeah, so think of it this way. In the old model, storage is so much passive, it's a place where you store the data. In the new world model, storage is actually active. It's active in moving the data and helping analyze the data, like for example, in that metadata example I just gave. So storage is not a passive device anymore, storage is an active element of the entire analytic machine learning artificial intelligence process, so you can get real insights. If you just relied on the CPU to do that, not going to happen. So the storage is now an active participant in this end-to-end solution that extends from on-premise into the cloud, as you guys have called it, the true private cloud, from Wikibon, the storage is active in that, versus being just a passive tool. Now it's very active, and the intelligence, and some of the things we've done with cognitive storage at the IBM site, allows the data, like our spectrum scale product, which is heavily involved in giant, hundreds of petabyte analytic workloads today, in production in major enterprises across the globe, as well as in high-performance compute environments, extending from on-premise out to cloud, but that storage is active, not passive as it was in the old days. So you mentioned cloud, so we're pretty strong believers in this notion of true private cloud, which is the idea that instead of thinking ultimately about in the industry that the architecture is going to move all the data to the cloud, that increasingly it's going to be moved cloud services down to the data and do things differently. And that seems to be picked, people seem to be, that seems to be resonating with folks. The question that I have then is, when we think about that, I've had, where's the data going to be located? It's going to have a major effect on where the workloads actually run. I've had three conversations with three different CIOs in the last six weeks, and they all said, I'm thinking differently. And instead of thinking about moving data up to the cloud, I'm now thinking about, how do I ensure that I always have control over my data, even if it's running in the cloud, because I'm afraid that if I move everything to the cloud, when I do have to bring it back, it's going to be such a huge capital expense that everybody's going to say no, and I can't do it. So it's almost like, maybe I'll do some stuff in the cloud, but I'll do backup, restore, or have protection on site. What do you think the role of storage is going to be as we think about multi-cloud and being able to do end-to-end, developing and putting various applications in various places? So you brought up a couple of topics there, right? So your concept and your research on true private cloud, actually, I find resonates amazingly well with clients. In fact, a lot of clients are trying to figure out a leveraged cloud, but they have a lot of data on premises and they want to leverage that. So the way I explain to clients, everyone wants to do everything they can do in the public cloud, all the agility, all the consumption models, all the DevOps models, and they just want to do the on-premises. So it's really an agility statement, but then extend to have the right workloads working the right hybrid cloud on their demand. But that brings a whole bunch of things. So the best use case I'll use, and then I'll get into the multi-cloud, but the one use case that all these companies, why did you end up going to Amazon or whatnot? And then it really gets down to developers. Developers were able to swipe or credit card, whatever, put their credentials in, swipe or credit card, do one line of code, spin up an environment, one line of code, spin down an environment, or they'd be a chef and puppet, and that would do the API calls, but they're able to do things very quickly. Try that in the enterprise. I mean, literally, they would have to go do a ticket, talk to Joe IT, which they don't want to do, it takes a lot of time, and it takes best case, about a week, four to five days, on worst case, up to three weeks to provision that environment. If you're doing agile development, it literally breaks the process of doing anything agile. So you're not going to do it. You're forced, you're absolutely forced to go away. So what we're doing is we're doing investments on-prem to do exactly, bring the agility. For example, the idea of swipe or credit card, we have a process, excuse me, a software product across as an API automation layer across all of our storage that gives you the last mile. How do you literally give API templates to your developers that they can literally, one line of code, spin it up, one line of code, spin it down, and it works across all our storage devices. But it took investment in another layer of API automation that the storage team sets up templates and able to, hey, gold, silver, bronze, provisioning your own storage, but in an enterprise way, or like a developer or an Oracle DBA, hey, spin up an environment free for test dev. But what we're able to do is a simple line of code to spin up a system, which should be, let's say four or five servers, last good snapshot from production that's been data-masked the way you need to do it, because you just don't give developers a whole database. But then literally that becomes a template that with role-based access, again, credentials, the developer or shuffer puppet natively can literally one line of code, spin up an environment and one line of code, spin it down. The benefit is on-premises you actually have your data. So unlike in Amazon, you're spinning things up, spinning things down, but it's not really running on what your production data looks like. You're literally able to keep that up to last night's data or the weekend before, but again, with all the data masking. But you can literally show, so our investment thesis is, we need to work on the next level of automation to allow people to truly do everything they can do in the public cloud on private, and we're making a lot of investments to do that. So it's actually one of our biggest investment thesis, and it really plays out well as far as clientele. You mentioned the next thing, and you can jump in on both of these, but you also mentioned the next thing is, well now, true private cloud allows you to easily extend to these different clouds. Well, then how do you keep track of where that is? How do you have, each one of the different clouds will have their own SLAs, but how do you manage it? How do you think through security? How do you know you're getting the right SLAs, and where do you put the right things for the right places? And one, there's management stacks that do that with software-defined storage, which all our products, all you do, we can run an extension of your device in any of the major public clouds and manage that securely. I can add a couple more, but do you want to jump in? Yeah, I think the key thing here is you've got to be able, in a true private cloud, the enterprise is mimicking what an Amazon or an IBM cloud division does, right? Except they're doing it in their own walls, on their own premises. Now that may be spread across the world if it's a global enterprise, but it's, if you will, it's their version of IBM cloud, but they want to be able to burst out. So all of our software-defined storage and even our array storage is designed so that if they need to move data from on-premise to IBM cloud, from on-premise to Azure, from on-premise to Amazon, they can transparently move that data. In fact, we can set up where they can automatically tier the data. When the data gets cold, boom, they dump it off to IBM cloud. That way the data that's in the private cloud, on-premises, if you will, but a private cloud that they configure is there for them to use and they take their access out for those, and by the way, talking to the chief security officer and the chief legal officer, they figure out what workloads is it okay to put out there at IBM cloud? And that way they have total control, but they have the flexibility of going out to the cloud, all done with the storage in an automated fashion. I think the key thing from a true private cloud perspective is storage as well as network and server infrastructure. They want it to be as automated as possible. They had the big town turn of 2008, yes, IT spend is back up, headcount is back up, but when you look inside the envelope of headcount, there aren't 40 storage guys at XYZ Global Enterprise. There's 20. They are now hired 40 people, right? So they got 40 people back, but the other 20 went to test and dev. They are not doing storage now. So those 20 guys need to be fully automated to support all these extra developers in a global enterprise and even small accounts now need that. So the true private cloud mimics IBM cloud, mimics Azure, mimics Amazon, and all those public cloud providers will tell you, they make their business by making sure it's automated, otherwise they won't make any money. So the private cloud does the same thing. And those 20 guys are now, as you said earlier, managing oceans of data where the business has no specific visibility and now that day is going to create value in the future. It's an extremely complex arena. So with that in mind, you guys have been invited to speak to the board of directors of one of the large enterprise clients about the value that storage will play in a digital business. What are some of the things that you tell them? So let me take that one first. So I think a couple of things. First of all, storage is not passive the way it used to be. You need to think of it as an active element in your cloud strategy to keep your data whole, to keep your data secure. And most importantly, to make sure your data offers value. So for example, you need to use all flash. Why? Well, because it needs to be instantaneous. It needs to connect right into that CPU as fast as possible to suck the data in so you can analyze it. And the guy who analyzed the data faster, for example, in dark trading and financials, if you're slower, you lose $10 million or $100 million. So storage is critical in that. So you want to A, let the board of directors know that storage is a critical component because it's not just passing it. Like we said before, it's active. So storage is an intelligence, not dumb. And people view storage historically as dumb. So storage is active, storage is intelligent. Storage is a critical element of your infrastructure, both in your private cloud, but also for what you do to cut costs when you do go to public cloud for certain workloads. And so you need to view storage as a more holistic part of how you handle your data, how you harvest the value of the oceans. Okay, if you're going to be fishing, you better make sure you get a lot of fish if you're going to feed the populace. And the more you do, I think of course you got to be able to protect it and you want to be able to secure everything. You can't do that if storage is as dumb and passive. So the board of directors need to see as data is your lifeblood, data is your gold. You have to mine that data and storage helps you do that. It's not just a place you stick it. It's not a vault to stick the gold in later. It's helping you mine the gold, refine the gold, get the value out of that gold. How do you do 24 carat versus 18 versus 14? What do you charge for that? Storage can actually help you do all that analysis because it's an active element. What would you say? I agree with everything you said. And I'll actually play it back to how you started this conversation, which is the digital business is he who uses his data. So I probably start there and I used a classic metaphor of your data is oil and he who refines it gets a value of it. And I agree it's not perfect metaphor, but it's really about getting insight and leveraging the insight. And that does translate to a couple of things. So it does matter that you have a secure, but also matters you have the right performance, either on premises or in the cloud and get the right insights. Typically the right insights is leveraging the data behind your firewall, which is your proprietary data, which is 80% of data in the world is just not available to a public search engine. It's behind the firewall. And by the way, when you're looking at your business, you might want to combine it with different things like we talked a lot about our Watson ability to, you know, let's say you're in healthcare and you can bring in oncology. So Watson oncology can help you with your data or the weather channel. We can bring in weather into a lot of different applications. So you want to leverage other data sets that are publicly available, but also your private data centers and get unique insights to it. And you want to work with someone that those insights are actually yours, which is really where IBM differentiates the cloud from everything else. So you want to bring in AI or cognitive, but we actually have cognitive based upon industry. We've actually trained, the thing between a cognitive and AI is actually that train cognitive actually has to learn. But once it learns is able to give you very interesting, you know, insights to your data. We do it by industry, which is a very compelling way to do a data. And the other thing is you want to protect your data either on-prem, it's not only protection as far as if you have a failure, come back up and running. So recovery, resiliency, but as much also insecurity. So you need to secure it throughout. And then the other thing I kind of highlight is more compliance and everyone doesn't want to talk about compliance, but the price of compliance is nothing compared to the price of if you get audited or you're, and you have to get compliance back and prove that. Just do it right from day one and you need to be looking data that you're doing on-premises or in the cloud, especially multi-cloud, you need to keep compliance and ownership of the data because it's a highly regulated environment and you're seeing new things coming out in Europe. You really need to be on top of it because the cost of that compliance, it might seem, geez, that seems like a lot, but it's nothing compared to if you, after a lawsuit or something, you have to come back from it. That's what I normally would talk to a board about. So had you been back at IBM or at IBM now for a while, it's about a year, you're by five quarters or so, something like that? Four quarters. Four quarters. And you've had a chance to look at the assets that IBM has, IBM's obviously being a leader and the tech industry's going to remain so for a long time. But what will IBM be as a leader in the storage industry? What does leadership mean to IBM? It's kind of the one IBM-specific question I'm asking. So I think it's important. What is IBM leadership going to be in storage? So I think, and maybe it comes to the hypothesis of why it came to IBM. You know, to be honest, I think IBM helps people get from where they are to where they want to get to and helps them do that in what I'll say is risk-reduced steps. But very few companies have the breadth of portfolio or capabilities like what we have in cloud and cognitive than IBM. I also think storage, as an industry, is going through a major change. It might be the next era is about data, but as far as the storage industry, it's in a lot of changes. So I think it's a, I use the term big boy game because it's not about doing the next array, which we do, it's as much applying the right analytics and understanding the true flow of data and the right security to do it effectively. When I looked at kind of IBM, I kind of did four things and I think it does play to where our vision is, right? I actually think it is changing and our clients are being disrupted and they're looking for a partner to help them. And it's not just disruption of technology or consolidation or price pressures, but they're being disrupted by these, the Uber of my business is XYZ is a verb. So I keep on saying that, but clients, every industry are getting disrupted. So if they're hesitant, if they're on their heels, they're not able to lean in on technologies the worst thing you can do. So what they need is a partner that knows and it can have the right vision and capabilities to lean forward and with confidence move forward. IBM has a history of going ear to ear with clients. That's the first thing and we calmly do it and clients trust that we know where to go in. And that's a lot to do with our primary research looking out there. Second thing, I think we have the right vision, the cloud and cognitive vision. No one argues with me. How do you get the insight or your data in that matter? It's your definition of a digital business is right on. He who use their data to their advantage is really a digital business and is an advantage by that. Three, it's broad portfolio. So storage, we have the, you know, broadest portfolio in the industry and you need that because as we help clients it's not helping them with the next storage ray. It's helping them, here's your business and it's different for everyone. Here's where you want to go to as far as your infrastructure and transformation and I help you get there over time. That takes a broad portfolio, not only in storage but also overall the right services, the right software analytics becomes a big thing where the number one company in analytics and that comes to bear for all our clients but also have the right services capability that's going forward. And then I actually think where IBM Storage allows you to lean in is really the biggest thing. We're going to help you simplify so you can lean in with confidence because that's what everyone's looking for, a partner to allow you to get there and very few companies are positioned as well as IBM Storage to do that. And now I know I'm taking credit for a lot of IBM pieces but that's the strength because that's the leverage of using an overall company to help you industry by industry with industry vertical knowledge really help you lean in with confidence so you can grow your business and transform. So let me build on that because at the end of the day your ability to make these kinds of commitments to your customers is a function of your ability to make these commitments to IBM and other IBMers history of keeping the commitments they make to each other. So IBM as a culture, and I've been around for a long time and worked with a lot of clients and these things up and down, good and bad at a product level but you're absolutely right. IBM has a track record of saying here's where we're going if you want to come with us we're going to get you there. And during periods of significant disruption that's not a bad type of partner to have. I'd use the term people kind of say sometimes it's trust. They trust us to get there. I think that trust is well placed. Again, I came from the outside a year ago. We're the last company with primary research, right? So you have to say where's it going? We actually do primary research to, there's a reason we'd be able to go ear to ear as a company for a hundred plus years is because we actually do that and then loud people go ear to ear. I know sometimes IBM downplays it, I actually think it's a strength. But you know, the Watson Research Center in many respects is creating the new errors and has for many years and is doing so today too. It helped clients through those errors without leaving you behind, which is something that's rare. You don't see it, our competitors don't have that. And I think that's a big thing. All right, so I'm going to close it here. Ed Walsh, GM of Storage at IBM, Eric Herzog runs product marketing for the storage group at IBM. I want to thank you very much for being part of this CUBE conversation. As we try to bring the experts that matter and they're going to have a consequential impact on how the industry evolves. Thank you very much for joining us for this Wikibon CUBE conversation. I'm Peter Burris, until we talk again.