 From the SiliconANGLE Media office in Boston, Massachusetts, it's theCUBE. Now, here are your hosts. Dave Vellante and Stu Miniman. Cloud is dead. It's all going to the edge. Where is it? Hi everybody, this is Dave Vellante and I'm here with Stu Miniman. Stu, where does this come from, this narrative that the cloud is over? Well, Dave, you know, clouds had a good run, right? It's been over a decade. Amazon's dominance in the marketplace, but Peter Levine from Andreessen Horowitz did an article where he said, cloud is dead, the edge is killing the dead, the edge is killing the cloud, and really we're talking about IoT and IoT's huge opportunity. Wikibond, Dave, we've been tracking for many years. We did the original forecast for the industrial internet and obviously there's going to be lots more devices at the edge, so huge opportunity, huge growth, intelligence all over the place, but in our viewpoint, Dave, it doesn't mean that cloud goes away. We've been talking about distributed architectures now for a long time. The cloud is really at the core of this. Building services that surround the globe live in just hundreds of places for all of these companies. So it's nuanced and just as the cloud didn't overnight kill the data center and lots of discussion as to what lives in the data center, the edge does not kill the cloud and we're seeing some major transitions pull and push from some of these technologies and a lot of challenges and lots to dig into. So I read Peter Levine's piece. I thought it was very thought provoking and quite well done. And of course he's coming at that from the standpoint of a venture capitalist, right? Do I want to start, do I want to pour money into the trend that is now the mainstream or do I want to get ahead of it? So I think that's what that was all about. But here's my question, Stu, is in your opinion, will the activity that occurs at the edge, will it actually drive more demand from the cloud? So today we're seeing the infrastructures of service businesses growing at what, 35%, 40%? Sure, sure. Amazon's growing at the 35% to 40%. Google, Microsoft are growing double that right now. So but overall you're right. Yeah, okay. And then of course the enterprise players are flat if they're lucky. So my question is, will the edge actually be a tailwind for the cloud in your opinion? Yeah, so first on your comment there, from an investment standpoint, totally can understand why edge is green field opportunity, lots of different places that I can place bets and probably can win as opposed to. If I think that today I'm going to compete against the hyperscale cloud guys, they're pouring $10 billion a year into their infrastructure. They have huge massive deployment so the bar to entry is a lot higher. I'm sorry, the second piece was. So will the edge drive more demand for the cloud? Yeah, absolutely. I think it does Dave, because let's take something like autonomous vehicles, something that we talk about. I need intelligence at the edge. I can't wait for some instruction to go back to the cloud before my Tesla plows into an individual. I need to know that it's there. But the models themselves, really I've got all the compute in the cloud. This is where I'm going to train all of my models, but I need to be able to update and push those to the edge. If I think about a lot of the industrial applications, flying a plane, things need to happen locally, but all the anomalies and new things that we run into, there's certain pieces that need to be updated to the cloud. So it's kind of a multi-layered if we look at how much data will there be at the edge? Well, there's probably going to be more data at the edge than there will be in the central cloud, but how much activity, how much compute do I need? How much things do I need to actually work on? The cloud is probably going to be that central computer still, and it's not just a computer. It's, as I said, a distributed architecture. That's where, when we've looked at big data in the early days, Dave, when we can put those data likes in the cloud, I've got thousands or millions of compute cycles that I can throw at this at such a lower price and use that there, as opposed to at the edge, especially what kind of connectivity do I have? Am I isolated from those other pieces? If you go back to my premise of we're building distributed architectures, the edge is still very early. How do I make sure I secure that? Do I have the network? There's lots of things that I'm going to build in a tiny little component and have that be there. And there's lots of hardware innovation going on at that edge, too. Okay, so let's talk about how this plays out a little bit. And you're talking about a distributed model and it's really, to me, a distributed data model. The research analysts at Wikibon have envisioned this three tier data model where you've got data at the edge, which you may or may not persist. You've got some kind of consolidation or aggregation layer where it's kind of in between the edge and the deep data center and then you've got the cloud. Now that cloud can be an on-prem cloud or it could be the public cloud. So that data model, how do you see that playing out with regard to the adoption of a cloud, the morphing of cloud and the edge and the traditional data center? Yeah, we've been talking about intelligent devices at the edge for a couple of decades now. I mean, I remember I built a house in like 1999 and the smart home was already something that people were talking about then. Today, great, I've got my Nest, if I have, I probably have smart assistance. There's a lot of things, I love that. So I'm on Twitter today, somebody's talking like, oh wait, I'm waiting for my light bulbs to update their firmware from the latest push. So some of it's coming, but it's just this slow gradual adoption. So there's the consumer piece and then there's the business aspect. So we are still really, really early in some of these exciting edge uses. Talk about the enterprise. They're all working on their strategy for how devices and how they're going to work through IoT, but this is not something that's going to happen overnight. It's, they're figuring out their partnerships, they're figuring out where they work and that three tiered model that you talked about is my cloud provider, absolutely hugely important for how I do that and I really see it, Dave, not as an or, but it's an and. So I need to understand, right, where I collect my data, where certain aspects are going to live and the public cloud players are spending a lot of time working on that intelligence layer. And Stu, I should mention, so far we're talking about really the infrastructure as a service layer, it comprises database and middleware, we haven't really addressed the SaaS space and we're not going to go deep into that, but just to say, I'm going to look packaged software as we knew it is dead, right? SaaS is where all the action is, it's the highest growth area, it's the highest value area, so we'll cover that in another segment. So we're really talking about that, the stack up to the middleware and the database and obviously the infrastructure as a service. So when you think about the players here, let's start with AWS. You've been to, I think every AWS reinvent maybe with the exception of one, you've seen the evolution, I was just down in DC the other day and they have this chart on the wall, which is their releases, their functional releases by year, it's just overwhelming what they've done. So they're obviously a leader, I saw a recent Gartner Magic Quadrant, it looked like, I tweeted, it looked like Ronnie Turcott, looking back on Secretariat from the Belmont and whatever it was, 1978, I think it was, 31 lengths. I mean, massive domination in the infrastructure as a service space. What do you see going on? Yeah, so Dave, absolutely. Today the cloud is it's Amazon's market out there. Interestingly, if you say, okay, what's some of the biggest threats in the infrastructure as a service? Well, maybe China, Dave, Alibaba's one that you look at there, but huge opportunity for what's happened at the edge. If you talk about intelligence, if you talk about AI, talk about machine learning, Google's actually the company that most people will talk about, it can kind of have a leadership. Heck, I've even seen discussion that maybe we need antitrust to look at Google because they're going to lock things up. They have Android, they have Google Home, they have all these various pieces, but we know, Dave, they are far behind Amazon in the public cloud market and Amazon has done a lot, especially over the last two years, you're right, I've been to every Amazon read vet except for the first one and the last two years really seen a maturation of that growth, not just devices and partnerships there, but how do they bring their intelligence and push that out to the edge? So things like their serverless technology, which is Lambda, they have Lambda green grass that they can put to the edge. Serverless is pervading all of their solutions, they've got like the Aurora database. And serverless is profound, not just from the standpoint of application development, but just an entire new business model is emerging on top of serverless and Lambda really started all that, but carry on. And when you look in and you say, okay, what better use case than IoT for, well, I need infrastructure, but I only need it when I need it and I wanna call it for when it's there. So that kind of model where I should be able to build by the microsecond and only use what I need, that's something that Amazon is at the forefront, clear leadership position there and they should be able to plug in and if they can extend that out to the edge, starting new partnerships, like the VMware partnerships interesting, Red Hat's another partnership they have with OpenShift to be able to get that out to more environments and Amazon has a tremendous ecosystem out there and absolutely is on their radar as to how they push out the IoT. So we were at Google next last year, big push verbally anyway to the enterprise and they've been making some progress, they're hiring a lot of people out of formerly Cisco, EMC folks that understand the enterprise, but beyond sort of the AI and sort of data analytics, what kind of progress has Google made relative to the leader? So in general, enterprise infrastructure service, they haven't made as much progress as most of us watching would expect them to make. But Dave, you mentioned something, data. I mean, at the center of everything we're talking about is the data. So in some ways is Google, come on Google, they're smarter than the rest of us. They're skating to where the puck is Dave and infrastructure services. Last decade's argument, if it's the data and the intelligence, Google's got just brilliant people, they're working at some of these amazing environments. You look at things like Google Spanner. This is distributed architecture. Say, how do I plug in all of these devices and help them work in a distributed gradual work? Well, heck, I'd be reading the white papers that Google's doing and understanding that they might be really well positioned in this 3D chess match that we're playing. Your eyes might bleed. I've read the Google Spanner. I was very excited about it and understood a little bit of it. Okay, let's talk about Microsoft. They're really of the big cloud guys. They're really the one that has a partnership strategy to do both on-prem and public cloud. What are your thoughts on that now that Azure Stack is starting to roll out with some key partners? Yeah, absolutely. It's the one that, Dave, if you use your analogy, looking back, it's like, well, the next one, it's gaining a little bit, gaining a little bit, but still far back there is Microsoft. Where Microsoft has done best, of course, is their portfolio of business applications that they have. They've really turned the green light on for enterprises to adopt SaaS with Office 365. Azure Stack, it's early days still, but companies, they use Microsoft, they trust Microsoft. Microsoft's done phenomenal working with developers over the last couple of years. Very prominent, like the Kubernetes shows that I've been attending recently. They've absolutely got to play for serverless that we were talking about. I'm not as up to speed as to where Microsoft sits for kind of the IoT edge discussions. You know, they're playing there. Absolutely, I mean, Microsoft's identity better than anyone, Active Directory is still the standard in enterprises today. So, I worry that Microsoft could be caught in the middle. If Google's making the play for what's next, Microsoft is still chasing a little bit what Amazon's already winning. Okay, and then we have enough time to really talk about China, you mentioned it before, Alibaba's legit 10 cent Baidu, obviously with their captive market in China, they're going to do a lot of business and they're going to move a lot of compute and storage and networking, but we maybe address that in another segment. I want to talk about the traditional enterprise players, Dell, EMC, IBM, HPE, Cisco, where do they stand? We talk a lot at Wikibon about true private cloud, the notion that you can't just stick all your data into the public cloud. Andy Jassy may disagree with that, but there are practical realities and you're certainly, when you talk to CIOs, they underscore that, but that notion of true private cloud hasn't allowed these companies to really grow. Now, of course, IBM and Oracle, I didn't mention Oracle, have a different strategy and Oracle strategy is even more different, but so let's sort of run through them. Let's take the arms dealers, Dell EMC, HPE, Cisco, maybe you've put Lenovo in there. What's their cloud strategy? Well, first of all, Dave, I think most of them, they went through a number of bumps along the road trying to figure out what their cloud strategy is. Most of them, especially, let's take the compute or server side of the business, they are suppliers to all the service providers trying to get into the hyperscalers. Most of them have, they all have some partnership with Microsoft, there's Azure Stack, and they're saying, okay, hey, if I want an HPE server in my own data center and in Azure, Microsoft's going to be happy to provide that for you, but David, it's not really competing against infrastructure as a service and the bigger question is, as that market has kind of flattened out and we kind of understand it, where is the opportunity for them in IoT? We saw, Dave, the last five years or so, can I have a consumer business and an enterprise business in the same? HPE tore those two apart. Michael Dell has kept them together. IBM spun off to Lenovo, everything that was on the more consumer side of the business. Where will they play or will companies like Google, like Apple, the ones that, Dave, they are spending huge amounts of money in chips. Look at Google and what they're doing with TPUs. Look at Apple, I believe it was, there was an Israeli company that they bought and they're making chips there. There's a different need at the edge and sure, a company like Dell can create that, but will they have the margin? Will they have the software? Will they have the ecosystem to be able to compete there? Cisco, I haven't seen on the compute side them going down that path, but I wasn't Cisco live and a big talk there. I really like the opening keynote and we had a sit down on theCUBE with the executive. It said, really, if I look out to like 2030, if Cisco's still successful and we're thinking about them, we don't think of them as a network company anymore. They are a software company and therefore, things like collaboration, things like how it's kind of a new version of networking that's not on ports and boxes, but really, as I think about my data, think about my privacy and security, Cisco absolutely has a play there. They've done some very large acquisitions in that space and they've got some deep expertise there. But again, Dell, HPE, Cisco, predominantly arms dealers, obviously don't have an HPE at one point, had a public cloud that pulled back. HPE's cloud play really is cloud technology partners that they acquired. That at least gives them a revenue stream into the cloud. Now, maybe it's- It's a consultancy. It's a consultancy. Now, maybe it's a one-way trip to the cloud, but I will say this about CTP. What it does is it gives HPE a footprint in that business and to the extent that they're a trusted service provider for companies trying to move into the cloud, they can maybe be in the cat bird seat for the on-prem business. But again, largely an arms dealer, it's going to be a lower margin business, certainly than IBM and Oracle, which have applications. They own their own public cloud with the Oracle public cloud and IBM cloud formerly software, which was a $2 billion acquisition several years ago. So those companies from a participation standpoint, even a tiny market share is compared to Amazon, Google, and Microsoft, they're at least in that cloud game and they're somewhat insulated from that disruption because of their software business and their large install base. Okay, I want to sort of end with the sort of where we started, the Peter Levine comment, cloud is dead, it's all going to the edge. I actually think the cloud era, it's kind of playing out as we, as many of us had expected over the last five years. You know what blew me away is Alexa. Who would have thought that Amazon would be a leader in this sort of natural language processing marketplace, right? You would have thought it would come from, so certainly Google with all the search capability. You would have thought Apple, but Siri compared to Alexa. So my point is Amazon is able to do that because it's got a data model. It's a data company. All these companies, including Apple, Google, Microsoft, Amazon, Facebook, the largest market cap companies in the world, they have data at the core. Data is foundational for those companies and that's why they are in such a good position to disrupt. So cloud, SaaS, mobile, social, big data, to me, Stu, these are kind of the last 10 years. The next 10 years are going to be about AI, you know, machine intelligence, deep learning, machine learning, cognitive. We're trying to even get the names right, but it starts with the data. So let me put forth the premise and get your commentary and tie it back into cloud. So the innovation in the next 10 years is going to come from data and to the extent that your data is not in silos, you're going to be in a much, much better position than if it is. Number two is your application of artificial intelligence. You know, whatever term you want to use, machine intelligence, et cetera. Data plus AI, plus I'll bring it back to cloud, cloud economics. If you don't have those cloud economics, then you're going to be at a disadvantage of innovation. So let's talk about what we mean by cloud economics. You're talking about the API economy, talking about global scale, always on, very importantly, something we've talked about for years, virtually zero marginal costs at volume, which you're never going to get on-prem. Because this creates a network effect. And the other thing it does from an innovation context, it attracts startups. Our startup saying, hey, I want to build on-prem. No, they're saying I want to build in the cloud. So it's data plus artificial intelligence plus cloud economics that's going to drive innovation in the next 10 years. What are your thoughts? Yeah, Dave, absolutely. Something I've been saying for the last couple of years. We watched kind of the customer flywheel that the public clouds have. Data is that next flywheel. So companies that can capture that. You mentioned Amazon and Alexa. One of the reasons that Amazon can basically sell that as a loss is lots of those people. They're all Amazon Prime customers and they're ordering more things from Amazon and they're getting so much data that drive all of those other services. Where is Amazon going to threaten in the future? Everywhere is basically what they see. The thing we didn't discuss there, Dave, I love your premise there is it's technology plus people. What's going to happen with jobs? You and I did the sessions with Andy McAfee and Eric Bronyolson. It's racing with the machine. Where is, we know that people plus machines always beat. So we spent the last five years talking about data scientists, the growth of developers and developers and the new king makers. So what are those new jobs? What are those new roles that are going to help build the solutions where people plus machine will win? And what is that kind of next generation of workforce going to look like? Well, I want to add to that, Stu. I'm glad you brought that up. So a friend of mine, David Michela, just is about to publish a new book called Seeing Digital. And in that book, I got an advanced copy. In there, he talks about companies that have data at their core and with human expertise around the data. But if you think about the vast majority of companies, it's human expertise and the data is kind of bolted on. And the data lives in silos. Those companies are in a much more vulnerable position in terms of being disrupted than the ones that have a data model that everybody has access to with human expertise around it. And so when you think about digital disruption, no industry is safe, in my opinion. And every industry has kind of its unique attributes, obviously publishing and books and music have disrupted very quickly. Insurance hasn't been disrupted. Banking hasn't been disrupted. A little blockchain is probably going to affect that. So again, coming back to this tail end premise is the next 10 years is going to be about that digital disruption. And it's real. It's not just a bunch of buzzwords. A cloud is obviously a key component, if not the key component of the underlying infrastructure with a lot of activity in terms of business models being built on top. All right, Stu, thank you for your perspectives. Thanks for covering this. We will be looking for this video, the outputs, the clips from that. Thanks for watching everybody. This is Dave Vellante with Stu Miniman. We'll see you next time.