 From the SiliconANGLE Media office in Boston, Massachusetts, it's theCUBE. Now, here are your hosts, Dave Vellante and Stu Miniman. Hi everybody, 10 years ago in 2006, Amazon introduced EC2, its first cloud service. And many of us at the time predicted that it would change the world. 10 years on, it has changed the world. The cloud has taken the IT industry by storm. We've been seeing this groundswell for a number of years now, and it seemed to hit in a big way, in a big spike in 2016, where major practitioners began to, at large organizations, began to look at the cloud, generally, in Amazon web services, specifically as a fundamental component of its cloud strategy, along with others, of course, including Google and Microsoft and many, many others. But the heart of this cloud trend really started 10 years ago with Amazon. And we're here to preview Amazon re-invent 2016. My co-host Stu Miniman, Stu, you've been all over this ecosystem for quite some time. This is probably the biggest show of the year. Give us the preview of what you expect in terms of attendance and action going on at the show. Yeah, Dave, it's a super exciting show. I know theCUBE's thrilled to be there again. It'll be theCUBE's fourth year, my fourth year at the show, and the fifth year for the show overall. And like public cloud itself, we've been seeing exponential growth. So what started out as a few thousand people in Vegas, the number I'm hearing for this year is going to be 25 to 30,000. And it was like between 18 and 19,000 last year. So just like revenues, Dave, the number of people coming is huge. It's going to burst beyond kind of the Venetian Palazzo. I hear there's some events happening at the win. People will be staying all over and we're going to have two and a half days of full coverage of theCUBE. John Furrier, myself. We're going to miss you there, Dave, but glad we can break down some of it here beforehand. Re-invent, it's going to be a while, I guess, before it outgrows Las Vegas with CES as probably a couple of hundred thousand people. But it is beginning to flood into other parts of the city and requiring more hotels. I've said, Stu, this is kind of, we're entering the fourth phase of cloud. The first one was kind of getting used to this notion of elastic capacity and pay by the drink. And a lot of startups, as we know, took advantage of that. The second great phase that we saw was when the economy turned down in 2008, 2009, CFO said, I want to shift CapEx to OpEx and people started to really pick up on that. When we came out of the recession, Shadow IT really started in a big way. And now you're seeing organizations bring back sort of cloud into the staple of the data center and CIOs really owning that. It's becoming a fundamental component, as I said in the open, of people's cloud strategies. Yeah, absolutely Dave. The first year we went to the show, there's all these just gamers, people doing mobile development, very much a developer centric attendee there. Everybody got footies back four years ago at the show. The last couple of years, it's that kind of back and forth with the enterprise. Last year felt like a lot of the enterprise arrived. There was a big ecosystem of partners, people that understand companies like Accenture, training hundreds of people to be able to deliver services to help people adopt public cloud. Lots of the other integrators and service companies are getting involved in what's happening there. And there's just such a swell of everybody being involved. When I talked to practitioners, there's two types of practitioners Dave, people that are using AWS and people that don't realize that they're using AWS. Because it's just everything from an hour when the quiver to the main way that people are doing business. Amazon's been talking for the last couple of years about being all in. As we poke and tease at that, we believe it's a multi-cloud world. Everybody, especially software as a service will be using services that sit on multiple clouds. But Amazon is the big one, biggest revenue still growing at a very rapid pace. And really impressive the way that they keep adding features, learning from where they need to go, listening to the customers. And as Andy Jassy talks about that flywheel of adding more customers and adding more features just keeps perpetuating itself. Well, let's break down some of the numbers. So Amazon began quantifying for the world its AWS financials in the first quarter of 2015. Yeah, and Dave, remember back then, everybody was like, wait, wait, Amazon, they're bleeding money, they're spending tons of money. The kind of common wisdom out there that kind of people thought wasn't what the reality was and wasn't what you thought it was. Well, you're right. I mean, I always had said Amazon is incredibly profitable. Once they break out the numbers, you're going to see that. And the reason is, as we've written about and talked about is that Amazon's volume and its efficiencies and its marginal economics at volume are so superior to its competitors. And so the incremental cost of Amazon deploying compute or storage or servers or databases, et cetera, goes toward, doesn't go to zero, but it goes toward lower and lower and lower each time. So its cost structure is incredibly efficient and attractive. So let's break out some of the numbers. So as I said, Q1-15 is when Amazon started to divulge some of its financials for AWS. 10% of the total Amazon company revenue was accounted for by AWS in Q4-2015. They hit a $10 billion run rate in Q1-16. And my Amazon model says they'll do just under $14 billion this year. We've got them coming in at $14.6 billion. I think that's somewhat of a conservative figure. They hit a $1 billion stew of operating income in Q3-2016. These are non-gap numbers. So as I say, I think they'll do just under $14 billion this year. $4.2 billion in operating income is the forecast that I have for Amazon. The most impressive part of Amazon, to me, and it relates to the marginal economics that I was talking about before, is in Q1-15, when Amazon first broke out AWS, it had 17% operating margins. I'm predicting that in Q4, it's going to top 32, 33% operating margins. So it's clearly into the mid-30s now, low to mid-30s. Just by point of comparison, EMC, an incredibly profitable company, its non-gap operating margins are in the mid-teens. Okay, so Amazon's operating at double the profitability of a company like EMC. That is really impressive. Yeah, absolutely, Dave. The two things I don't hear anymore. Number one, the discussion of race to zero seems to have gone away. We understand that there will be price decreases as these cloud services mature, and as there's more competition, Microsoft's coming on hard, Google with Diane Greene making a big push here, but it is no longer a discussion of, oh, this is just, I mean, some vendors still say, there's this commodity cloud out there that's just going to be the race to zero, but if you look at the numbers, as you said, they're very profitable, very impressive what they're doing, and yeah, it's just quite impressive what they're doing. Some of the discussions that we're having in the Wikibon community relate to what's happening on prem. We saw the AWS VMware announcement earlier this year. Many thought, again, the discussion in the communities, there's pluses and there's minuses. Many see that as a win for VMware because they can now offer cloud services from AWS and extend their franchise into the cloud. Others see it as a one-way trip to the cloud. So there's some interesting debates there. There's also a debate going on internally as to what Amazon will do about the Azure stack, AWS specifically. David Floyer has predicted that Amazon will do some type of on-prem Azure stack-like appliance. You have said, you doubt it, you're a real skeptic in that regard. Maybe you talk about that a little bit. Yeah, sure. We've seen a lot of services from Amazon helping to migrate to the public cloud. One of the lines that you talk about VMware on AWS is we've just stuck a straw into 500,000 customers that Amazon is going to just kind of pull in all of those workloads. And all the conversations we've had with Andy Jassy and looking at Amazon, it's hybrid cloud, they think is a halfway house. They want everything to go in the public cloud. Now that being said, I understand from an architectural standpoint, David Floyer believes that IoT requires edge computing and things need to happen there. That's where I'm not opposed to seeing that Amazon would have some device that would do some kind of processing out of the edge and only move the data that needs to be moved because just the pure physics of moving all the data from the edge to the core probably doesn't make sense. But when I look at Azure stack, Azure stack is like a, you know, infrastructure replacement. It's here's your, you know, Dell EMC or HPE or Lenovo hardware stack. And it looks more like traditional infrastructure. And we're just going to mirror it between the public cloud and on-prem. I don't expect to see an Amazon stack that's going to be used to kind of replace your sand because Amazon believes that, you know, where their services are and, you know, all the activity that they have are in the public cloud. But by the way, the other thing that I had meant to mention before is for a while people were thinking that maybe AWS should be spun out of Amazon. And I haven't heard that for a while. It's the second thing that kind of seems to have fallen by the wayside lately. Well, it surprised me for AWS. I mean, it's a large proportion of Amazon's operating profits. It throws off cash. It just seems in a weird way. It is synergistic because you've got the retail operation heavily using, you know, AWS. That's how that was those are the roots of AWS. You know, it's also interesting. When you think about competing with Amazon and the knockoffs on Amazon, it used to be SLA, SLA, SLA. That's you don't hear that much anymore. And then it used to be, well, they don't have a lot of data centers. You're now seeing Amazon extend its data center footprint into places where there are, you know, local laws requiring data. But where are people able to compete with Amazon? I mean, obviously Microsoft is having some success with Office 365. They're taking a different approach from a partnership standpoint with the Azure stack. But what are you saying? Yes, so first of all, right, Dave, start with the applications. You know, Amazon has a number of services that they build in and certain SaaS providers that will live on Amazon. But if I can build a service that meets what my customers need, you know, I'm going to use that SaaS offering and I don't really care as much what's sitting underneath it. Secondly, it's really around kind of the governance and compliance issue, Dave. So absolutely Amazon is growing their footprint. They're growing, grow globally. But there are people that are going to want to, you know, hold it in-house. You know, the financial community is still very much wants it within their four walls or they're working with service providers that are very local. And I was just over in Austria recently and talked to, you know, a service provider in South Africa, another one in the Netherlands. And, you know, on a day-to-day basis they are not competing with Amazon. You know, one of them focuses heavily on retail. Of course, the whole retail industry is probably not going to go on ADBUS because that is, you know, competitor number one. Walmart, you know, doesn't use AWS because Amazon.com is their chief competitor. So there are definitely certain verticals that are leveraging public cloud quite strongly. But, you know, service providers and other services that can get local and handle that governance and compliance issue. That's something that's still going to take a long time to sort out. The other thing, and I heard Roger McDonough be talking about this the other day, that Amazon for the first time since, you know, Windows, you're seeing software companies, ISVs placing their product on Amazon, running it in Amazon. And of course, he pointed out, you know, you saw what happened with Microsoft. They just sort of gobbled up the application space. You're seeing Amazon, and Roger didn't reference this, but Amazon's already moving up the stack with database. With, you know, DynamoDB, DB, Aurora. I mean, they've got a number of, you know, Redshift, RDS. They've got a number of database offerings and middleware offerings. They are moving up the stack. Do you expect, is there a discussion going on about Amazon eventually getting into the, you know, the applications business? There's a ways off, but at some point, when they get to be $50 billion, they're going to be needing to expand their TAM. What are your thoughts? Dave, every year at Amazon, it's like, okay, what part of the ecosystem is Amazon going to take a big chunk out of? What was it? It was, you know, a year or two ago, it was the database stuff, you know, focusing on what Oracle's doing with Aurora. It was Citrix a few years back when it was the, you know, workspaces that was going on there. With the VDI. Absolutely. I mean, Dave, the word when we talked about Amazon is frictionless. How does Amazon make it easier? They had a recent announcement for SaaS providers that are in the marketplace. I can actually just pay all that through my Amazon subscription. No longer do I have to say, oh, okay, wait, I'm going to use all of these services and manage this billing here and this billing there. I could just wrap it all in. Amazon's going to be a single bill, put it on, you know, a single contract, make it nice and easy for me to use, you know, SaaS providers that are in the Amazon marketplace, Amazon services, and it just becomes so easy to be my supplier of choice. Yeah, and think about a total available market standpoint. I mean, I think, I mean, I've talked to Andy Jassy about this before and I've said it's a trillion dollar TAM. I don't know if he's used that number, but I have no doubt that they're looking at that, you know, large space. And so there's another dynamic here, Stu, and we're seeing the, everybody's an IT company now. Banyoff said it, I think first, is that you're going to see more SaaS companies come out of non-technology companies than you will out of technology companies. And that's to me really the dynamic that Amazon is playing. They're enabling non-tech companies to become tech companies, taking the infrastructure off the table and saying, okay, you go and innovate. Those practitioners are not threatened by Amazon. You know, data stacks, putting its database on Amazon, that might be a threat there because Amazon might be pushing DynamoDB. And so you're going to see some friction, I would predict, within the IT supplier space. But beyond that, the practitioner space, I think that's where the huge opportunity is. Yeah, Dave, absolutely. I mean, it's why you saw companies like Nike, like Capital One, others getting up on stage, talking about how they're getting rented data centers, standardizing on services from Amazon. It's interesting, GE has been up on stage there. Predix, I believe, lives on Azure though. So this dynamic of where resources live and public cloud, just being that API economy, software services is public cloud where a lot of that's going to live. Well, it's a multi-cloud world, but I want to come back to the comment you made on API. So there is also a discussion going on in the Wikibon community about API creep and API complexity. Amazon strategy has always been get down to the API level. That's the primitive construct and it allows you maximum flexibility. If you go too high of an abstraction layer, you make mistakes and then you're sort of locked into that abstraction layer and you can't be agile and you can't be flexible. So Amazon's choice seems to anyway, Stu, I'd love your thoughts on this to be, to go down to the granular API level. But if you look at, you know, Amazon's got at least 10 data services with EC2 and S3 and Glacier and Kinesis and DynamoDB and on and on and on. Each of those has a separate proprietary API. And so the data pipeline starts to get elongated and you're moving around this set of different services. It starts to get complex. Also, you're making choices based on cost and based on the frequency of access of your data. And a lot of times it's not clear. Sure, Glacier's clear. I'm going to put it there and hopefully I never have to get it. But, you know, DynamoDB versus Redshift versus, you know, Aurora and Kinesis and it starts to get somewhat confusing for customers and they get that Amazon bill at the end of the month and they go, whoa. And then they got to move and so, or do something. So there's a bit of a backlash, is there not? Yeah, Dave, you know, I remember back, you know, five years ago we talked about kind of the hyperscale guys versus the enterprise guys. The enterprise guys are usually willing to spend money to have a solution that's nice and baked because I don't have the resources to kind of put all the pieces together. Amazon's a weird one because I remember the first year I went to Amazon, they came out with a new service that was like, we're going to give you better visibility into what you're doing and show you places where you've been like overspending like by 50 or 70%. And it shocked me coming from the enterprise space to hear people that were like, oh, this is a wonderful tool and oh, cool. I can save like half the money now that I was saving before. If this was the enterprise and you said, hey, you've been using this thing for two years and sorry, you've been overpaying by 50 to 75% for the last two years, they would have killed you and gotten rid of you. So it's this weird dynamic. Amazon looks at it as you said, Dave, you know, that API creep, you know, it's okay that I've got, you know, well, there's a thousand services out there. I'm going to choose the ones I have and I'm going to put it together. And there's lots of opportunity for those integrators and those service providers to help me, you know, leverage it. If there's something that I want, chances are it's going to be there. So it's the kind of, you know, big, you know, fries electronics that I walk in and I can put all the pieces together. But it is this dynamic, lots of opportunity for those kind of cloud management platform solutions out there and starting to see a number of companies that are trying to help solve this complexity problem, which I thought going to public cloud was supposed to solve that, right? We get this, you know, homogeneous environment, but, you know, enterprise, I've got lots of applications. I need lots of services and therefore, you know, I don't necessarily get rid of complexity, I've just kind of shipped it. Well, it hasn't seemed to hurt Amazon. I mean, you know, Jassy always points out that it's not about the costs, even though they claim, you know, they do lower their costs, but everybody lowers the cost every quarter. He used to work at EMC and the sales guys would quote lower price per gigabyte, for instance, every quarter. That's normal because of Moore's law, but it really, the point being, people don't go to Amazon just for the cost or even necessarily for the cost, they go for the flexibility, the agility. So my point is it hasn't hurt their growth rate. I mean, if you look at Amazon's growth rates, it's 50%, 80%, 78%, 55%, 64%. I mean, these are the annual growth rates of AWS services. They've stayed consistently above that 50% level. You know, people are waiting for it to attenuate. I mean, you saw that at Nutanix.next, D. Raj's presentation, he sort of gave a law of large numbers and showed Amazon tailing off. That hasn't happened yet. You know, we'll see your thoughts on their ability to continue growth. Can Amazon keep growing at this pace, AWS? Yeah, I think they can, Dave, because first of all, you talked about that TAM. And even I heard people this year say, well, Microsoft's growing at 100%. But, you know, from a public cloud and that's kind of SaaS and Azure all put together. But if you look at, you know, Microsoft and Amazon, absolutely they're competing, but Microsoft has their set of applications. They've got like Citrix applications living on top of them. Amazon is a lot of new applications. Communities I look at are super excited about things like AWS Lambda, which is the poster child for the serverless kind of functions as a service movement that's going on to really be able to take the application at the center using things like microservices and containers and not have to worry about the infrastructure underneath it. Just going back to that simplicity that we're trying to drive at. But I don't see anything that's gonna stop Amazon for at least the next few years. They've still got lots of global growth and tons of applications and pieces. And when you look at that overall TAM, there's still a small piece of it, even if they hit that $14 to $15 billion this year that puts them, where does that put them compared to some of the infrastructure guys, Dave? Well, I mean, so again, well, look at Dell EMC, with 80 billion, 70, 80, 85 billion. So ways to go there. So I think you're right. I think there's a huge enough market for Amazon to continue to grow. And I think the other thing is it's kind of bromide when you hear Jassy talk about how, well, we focus on the customers, but they do focus on the customers. It's the customer first. And they seem to be doing a pretty good job of listening to customers and delivering services that are in demand because they launched these new services and people eat them up and they grow quickly and they get testimonials and in real world, you know, large organizations. So that's the other big thing. You're starting to see Amazon really knock down some pretty impressive logos in the enterprise. They still haven't knocked down the hardcore, you know, Oracle workloads. I'm not saying they're not running Oracle on Amazon, people are, but the hardcore, you know, rack workloads, that's sort of Oracle sweet spot still. And that seems to be where Oracle is sort of migrating. Well, let's close on sort of the horses on the track. You got Amazon, we talked about how they're, you know, really great momentum. We've talked about Microsoft. Where's Google and all this? Yeah, so Google absolutely is trying to focus on the enterprise more, Dave. That's why they brought Diane Greene into the mix there. They've got a strong position really in the analytics space. They're going to be holding their cloud event in March and they've been growing, but they're kind of a far third in the big public cloud guys behind Amazon and Microsoft today. You know, they've been making some acquisitions. They're making some moves, but, you know, I don't hear Google as much yet, even though it's funny, some of the smart guys I know in the industry, they're like, oh, Amazon came out with a new feature. You know, Google had that two years ago. I said, yeah, but the knock on Google is always that Google builds something that Google engineers can build. You know, we just had, John Furrier was out at KubeCon. Kubernetes, of course, came out of Google. That's a way to get kind of that cluster layer of applications into all of the environments. So it's somewhere that kind of Google technology is going to spread quite a bit. So, you know, Google's definitely one to watch and, you know, how it plays into lots of partnerships that it has with, you know, companies like VMware. You know, Pivotal is now doing, I think it was Home Depot based on the Google Cloud Platform. And Oracle obviously seems to be focused on preserving its estate, its database franchise and then moving things into applications. I think Oracle sees that applications are really the next control point. And it also, I think Safra Katz has pointed this out many times, if you're selling the infrastructure and the middleware and the database and the applications, you're actually going to make more money in the long run. Oracle, obviously, very financially savvy. The other one's IBM. IBM, to me, it's one word, Watson. You can't get Watson without the soft layer cloud or whatever they're calling the cloud these days. So that's their strategy. Yeah, I'm interested to hear kind of discussion of machine learning and artificial intelligence. I'm sure it will come up at AWS this year. And we'll see how much of that competes against the Watson type of solution. Oh, it was interesting to see that 60 minutes interview with Charlie Rose several weeks ago with John Kelly and a bunch of others at IBM. Amazon, Microsoft, Google got virtually no love in that interview. And a lot of people say that those companies have better technology than Watson. So we'll see. But IBM from a marketing standpoint and a services standpoint, and it's got an early lead there, particularly in the enterprise. It doesn't have, in my view, anyway, hardcore competition in the enterprise. Somebody tweeted me the other day, well, what about Microsoft? You know, Microsoft, it's early days, they're still in beta. Not that Watson's not early days, but IBM has been doing some heavy lifting in certain industries like healthcare and certain other sort of peripheral industries. So I think it's got a head start there, but the key, again, to me for Amazon or IBM's cloud is it's got Watson. That's its sort of secret sauce right now. And it can sort of match the capabilities like it goes out and buys Cleversafe or ObjectStore, can match those capabilities that Amazon has. But Watson is really the future of IBM. All right, I'll give you the last word, closing thoughts. We're going to be there with theCUBE. What do you expect? Yeah, so Amazon, Dave, this is the show. I tell you, people ask me what shows this you go to. This is a show to be at. Just the people that are there, the customers that are at this. First of all, not only do you have good developers that are there, but the community of people that are there, they're excited about trying different things. And that's like, if you talk about companies and how you get ahead and stay ahead of your competitors, that being willing to try, risk, and take advantage of some of those new things out there, this is probably the best show out there. And now it's also one of the biggest shows out there. So super excited to be there. There's tons of social activity. I mean, I'm disappointed. I'm going to be so busy doing theCUBE. I can't do things like check out the preview of season two of Man in the High Castle, which is coming out on Amazon streaming. Dave, I know both of us were prime members. Amazon keeps adding all of these services that there's things going on there. So evening events, kind of the Cloudorati community that I like to hang out with some will be there. A bunch of the BCs that used to have parties at BM World are now having them at Amazon this year. So great group of people, super excited. I know sessions are going to be overloaded. The show floor is going to be nuts. And, you know, boy, Vegas is going to be packed with a ton of people excited about this. And we'll be right in the center of it on the show floor. Yeah, huge set there as always. You know, Amazon always gives us great space and we have awesome guests. We didn't talk about IoT, but check out some of the stuff on wikibon.com. We've been talking a lot about the Cloud and IoT and Edge and where Amazon fits in there. It'd be interesting to see what comes out of reinvent with IoT, check out siliconangle.com for all the news and siliconangle.tv for all the videos that we do and all the cube action. So thanks for watching everybody. This is Dave Vellante and Stu Miniman. We'll see you next time.