 Live from Las Vegas, it's theCUBE, covering AWS re-invent 2018. Brought to you by Amazon Web Services, Intel, and their ecosystem partners. Okay, welcome back everyone. Day three, we're live in Las Vegas for AWS re-invent 2018. So our six year covering Amazon re-invent and AWS, Amazon Web Services, a meteoric rise in value, profitability, market share, just a rising tide floating all boats. I'm here with Dave Vellante for kicking off day three analyzing, you know, Werner's keynote. Things start to wind down yesterday. It was kind of the big day with Andy Jassy, Dave. After yesterday, it's pretty clear that there's a couple of big mega trends that people are talking about. One, AWS Outpost, okay? That is going to be a one year conversation about what that means, what implications. I mean, basically, if you're a cloud native company, you order a data center from Amazon Primal deliver it in two days. Why would anyone want to buy hardware again from HPE or other companies? This is a huge risk, huge challenge, a huge shot across the battle of the industry because this is essentially saying, this is essentially cloud in a box, put it in, plug it in, the service turn it on and it works and developers just do their thing. That's amazing. So I think that's going to be a very hotly contested topic throughout their last, at least one year until they ship that and all the posturing and jockeying is going to go on there. And then the other thing that was interesting was it was a lot of coolness. The F1 racing car with analytics, you had Lockheed Martin with space satellite provisioning. That was pretty cool. And you got robots and IoT. That's cool. You got space, you got robots, you got sports cars, all using analytics, all using AI, all using large scale compute storage and networking, very elastic, all with all kinds of new tools and reference engines. But Jerry Chen laid it out from Greylock yesterday around the strategy. Amazon drive this cost down on the infrastructure side and bring the API concept up to AI and bring the marketplace together. So a lot of action. Today we're going to see the impact and the fallout of that. What's your thoughts? Well, first of all, John, there's so much to talk about. I want to say, so Warner Vogels this morning gave the keynote. When I first joined this industry, IBM was everything. IBM was the dominant player. So we used to pour over IBM system and technology guides and IBM white papers because they set the technical standard for the industry and they shared that knowledge obviously with their customers to inspire them to buy more stuff. But they were given back to the community as well to help people understand architectures and core computer science. Listening to Warner Vogels today, Amazon is now the beacon of technology in the industry. He went through the worst day of his life, which is December 12, 2004, when their Oracle database went down for 12 hours because of a bug in the code and because they were pushing it beyond its limits. And so he described how they solved that problem over a multi-year effort and really got heavy into the technology of database and recovery. And it was actually quite fascinating. But my takeaway was Amazon is now the company that is setting the technical direction of the industry for the next wave of cloud-based applications. So that was actually really fascinating. We heard similar things on S3 and S3 Recovery. Even though they're still using some Oracle stuff, it was really, really fascinating to see and very, very impressive. So that's one. As you say, there's so much to talk about. The IoT pieces, John, I really like what Amazon is doing with IoT. They're coming at it from a bottoms-up approach. What do I mean by that? Do you remember when mobile first came out? Microsoft basically said, hey, we're going to put Windows on a phone top-down. We're going to take our existing IT desktop standards. We're going to push them down to mobile. Didn't work. And I see a lot of IT companies trying to do that with IoT today, not Amazon. Amazon's saying, look, we're going to go bottoms-up and serve the operations technology people with a software development platform that's secure, that allows it, that's fully managed and allows them to build applications for IoT. I think it's the right approach. I think the other thing that's coming out is a tweet here from Bobby Allen, who we know from theCUBE days. I shared a tweet about the future of the converged infrastructure and the outpost. He says, software should be where tech companies differentiation lies, value lies. This is back to our beating of the drum about software, software, software. Andy Bethelstein, the Rembrandt of motherboards, Pat Gelsinger calls them, said he's the founder of Arista. Hardware's easy, software's hard. Software is where the action is. What Amazon's doing is essentially pushing large-scale platform capabilities and trying to make that as cheap and affordable as possible with a range of services while creating a new shim layer around API concepts and microservices up the stack to enable people to write software faster, more compelling, more meaningful, and to iterate. And this is resonating with customers, Dave, because if I'm a business, I got to write software, okay? I don't want to be in the running data center business because the data center powers the business. So the end doesn't justify the means in that regard. You say, hey, I need the data center to power my top-line revenue, which is either going to be software-based or some sort of edge network scenario or even a human interface wearable or whatever. Software is the key. So if Amazon can continue to push the cost structure, the lock-in spec is locked in because of the better value. So if it's going to be 80% less cost and you call that a lock-in spec, I call it a lock-in spec. It's not like a technical lock-in spec. That's just called value. I'm locked into Google search. I mean, I don't have to tell you. I'm not going to use any alternative search. I just don't know if I'm familiar with it. Better, I like it. It's better. But software is the key, your thoughts. Okay, so, well, my thoughts on lock-in are, lock-in is one of the most overstated concepts in the business. I'm not saying that lock-in doesn't happen. It does happen. It happens everywhere. It happens across open source. You do open source, you're locked into your developers. I've done research on this, John. And my research shows that 15% of the buyers really make primary decisions based on whether or not they're going to be locked in. 85% look at the business value and they trade that off against lock-in. So, you know, yeah, buyer beware, blah, blah, blah. But I think it's just really overstated here. Yes, it's the cloud, mother of all lock-ins. But what's the value that you get out of it? Speaking about another lock-in, I want to talk about Intel a little bit because the press has been like chirping about Intel and alternative processors and the ARM-based stuff that Amazon is doing. Well, hold on. Let's just set the table on this conversation. Intel announced a series of proprietary processors, their own silicon. You mean Amazon, you mean? I mean Amazon. Yeah, some proprietary processors that are specific to certain workloads, inference engines, and other things based on the Anapurna acquisition of 2015, a smaller Israeli-based company that they acquired. So the press, I've been chirping on, oh, chips must be confronting Intel, your thoughts. Yeah, so here's the deal, look it. Intel is massive and they do a huge amount of business with the cloud players. Now, here's the thing about Intel. That's really, I've observed Intel for decades. Intel wants a level playing field amongst its customer base and so it wants a lot of different cloud suppliers even though there's three, four, five worldwide. There's many dozens and hundreds of cloud players out there. Intel wants to support them all. They're an arms dealer, right? They love all their customers and so what they do is they sprinkle around the innovation and the industry. They try to open up their architecture such that people can write software to their architecture and they try to support all the customers. We see it all the shows. You see it at Lenovo, you see it at Dell, you see it here at AWS, you see it at Google. Intel is everywhere and they are by far the biggest supplier. Now, Amazon, of course, has to have alternatives, right? They care about data center power. They do buy some stuff from AMD. Why not? Why wouldn't you second-source some of this stuff? They do a lot of work with Nvidia. ARM has its place and so, but it's a rounding error in the grand scheme of the market. Now, why people get excited is they say, okay, ARM now has a foot in the door. Ooh, Intel's in trouble. Intel, obviously, still a dominant player. I think it's, you know, it's Intel in trouble. Press likes to glom onto that as Intel's like the dominant player in the microprocessor business and it has to move and it has to move fast. I would not say Intel is in trouble. I'd say it continues to be the dominant player in the data center. It's got opportunities for alternative processors like, Intel's strategy is to put as much function on the die as possible and to grab that function. It's always the way it's behaved. You see people like Nvidia trying to create opportunities and doing a very good job of it. And so, there's white space there. It's competition. We love competition, right? Intel needs some competition, frankly. Here's my take. One, Intel pays, Amazon pays Intel a lot of money. Huge amount of money. So it's not like Intel's hurting. Intel's not in trouble. Here's why Intel's not in trouble. One, the cloud service provider, BusinessEd, Regine runs, she was on theCUBE yesterday, is growing significantly. A new total addressable market called TAM expansion is happening. So, if you look at microprocessors, it's not a one or few suppliers. It's a total TAM expansion. And of course, with that expansion of the market, Intel's going to take a big chunk of the share. So they are not in trouble. Amazon pays them a lot of money. They're a big time supplier to AWS. Check. Two, Intel is on a cadence on processor design that spans years. And Regine and other Intel executives have spoken to us off the record and here on theCUBE that hey, you know, sometimes just use cases where they're not responding fast enough that are outside their operating cycles. But as Regine said, Amazon makes them get better, okay? So they have to manage that. But there's no way Intel's in trouble. I think the press are using this as to create link bait for news that is sensational. But yeah, I mean, on the surface you go, oh, chip, Intel, oh, that's Intel's business. It must be bad for Intel. So yeah, Amazon made their own processor. They got some specific things they want to build, specialize processors for, like GPU alternative or inference engines that are tied to the stack. Yeah, I wouldn't dig. What Intel will do, what Intel will do is they'll learn from that and they'll respond with functionality for maybe others or maybe they'll earn Amazon's business, we'll see. But to your point, you know, Intel's exposure to the desktop and the laptop, a lot of people wrote about that. And Intel has, the barriers to entry into Intel's business are so huge, the cost of doing what they do. Intel's such a strategic supplier to so many companies. And as we talked to Regine about yesterday, the cloud has completely changed that dynamic and actually brought more suppliers. The data center consolidation that you've seen has been offset by the cloud explosions. That's a good trend for Intel. And of course the mobile dynamic, you know more about that than I do, but everybody said mobile's going to kill Intel. It obviously didn't happen. Look at Intel, Intel's smart. They've been around. They're going to not miss the ball. They got a big team that services a lot of these big players. Are they still paranoid in your opinion? I think they are. I do too. I do too. I mean, look at Intel is have a cadence of Moore's law. They have a execution style that's somewhat similar to AWS. They're very strict about how they execute and they have a great execution engine. So I would bet the farm that Intel's talking to Amazon and saying, what do you need for us to be better? And if Amazon does what they do best, which is tell them what they need, Intel will deliver. So I'm kind of not worried about Intel on that front. I think in the short term, maybe this process doesn't fit for that, but that's why GPUs became popular. Floating point was a unique thing that CPUs didn't do well on. So a GPU comes out. There it is. We're going to see processors like data processing units, Pradeep Sindhu, former founder of Juniper's got a venture called Fungible that's building a data processing unit. It's a dedicated chip to serve analytic workloads. These are specialized silicon chips that are going to come on faster and to the marketplace. So just because there's more chips doesn't mean Intel dies because if the TAM expands, it's overall bigger market. So their share might not be as dominant on a smaller market, but it still wins. I got to come back to your John Chambers interview. I've watched it a couple of times now I would recommend people go to thecube.net and see John Furrier's interview with John Chambers. The great companies of this industry have survived, I talked about Paranoia, Andy Grove. They've survived because they were not dogmatic about the past. So for the past several decades, this industry has marched to the cadence of Moore's law. And that was obviously very favorable to Intel. Well, that's changing and that's changed. The innovation engine now, you've called it the innovation sandwich, is data, machine intelligence applied to that data and the scale of cloud. So Intel has to pivot to that to take advantage of that. And that's exactly what they're doing. So the great companies of the future, the Microsofts, the Intel's, the AWS's, they survived because they can evolve. It's the wangs that didn't, they denied. It was the PC, they were entitled, digital, right. They thought they were entitled. And the point that John Chambers made is there's no entitlement. And he kept referring to Boston 128. It used to be the Silicon Valley. And the leading executives today of companies like Cisco, like Intel, like Microsoft, can see a vision to the future and they change when they have to change. So companies that are entitled, who are they? Wow, that's a really good question. Is Oracle entitled, HPE, Dell? I think Oracle absolutely acts as though they're entitled and they're bunkering down into their red stack. Now, you know, I've often said, don't bet against Larry Ellison and I wouldn't make that bet against Larry Ellison, but his tam is confined to Oracle customers. He's not currently going after non-Oracle customers, in my opinion, at least not with a strategy that's obvious to me. And I think that's part of the reason why Thomas Currian left the company is I think they had a battle about that. At least that's what my sources tell me. I haven't talked to him directly. I actually don't know him, but I know people who know him and have worked with him. HPE, I think HPE is more confused as to what the next step is. When they split the company apart, they kind of gave up on software. They gave up on an integrated supply chain. Michael Dell took the other approach and thanks to VMware, he's got a winning strategy. So I think today's leading executives realize that they have to change. Look at Ginny Rametti. IBM was in trouble in my opinion because Watson failed and their cloud strategy essentially failed. So they just made a $34 billion acquisition, a Red Hat, which is a bold move. And that, again, demonstrate a company who said, okay, hey, it's not working, we have to pivot and we have to invest and go forward. All right, Dave, great kickoff day three at Andy Jassy coming up at the end of the day. And he's going to do his annual kind of, end of the last day, round up on theCUBE, kind of lean back, talk about what's going on and how he feels and the quotes, what people missed, what people got, and do a full review of re-invent 2018. Day three kicks off here, CUBE. Two sets on the floor getting all the content. We already have over 100 videos. We'll have 500 total video assets. Go to SiliconANGLE.com and check out the blog there. A lot of stories flowing, a lot of flow, a lot of demand for the content. Stay with us for more. Enter this short break.