 Hello everyone, welcome back to live coverage day two, or day one, day two for theCUBE, day one for the event. I'm John Furrier, host of theCUBE. It's the keynote analysis segment. Adam Slutty's just finished coming off stage. I'm here with Dave O'Luffin and Zia Carvella with the principal analyst at CK Research. Zia, it's great to see you, Dave. Guys, the analysis is clear. AWS is going next gen. You guys had a multi-day analyst sessions in on the pre-briefs. We heard the keynote, it's out there. Adam's kind of getting his seed legs, so to speak. A lot of metaphors around ocean, space, he's got these thematic explorations. He chunked his keynote out into sections. Zia, there's a lot of networking in there in terms of some of the price performance, specialized instances around compute, this end-to-end data services. Dave, you were all over this data aspect going into the keynote. Obviously, we kind of had visibility into this business transformation theme. What's your analysis? As Zia, we'll start with you. What's your take on what Amazon web services is doing this year and the keynote, what's your analysis? Well, I think there was a few key themes here. The first one is I do think we're seeing better integration across the AWS portfolio. Historically, AWS makes a lot of stuff and it's not always been easy to use, say, Aurora and Redshift together, although most customers buy them together, right? So they announced the integration of that. It's a lot tighter now. It's almost like it could be one product, but I know they like to keep the product development separately. Also, I think we're seeing a real legitimization of AWS in a bunch of areas where people said it wasn't possible before. Last year, NASDAQ said they're running in the cloud. The options exchange today announced that they're going to be moving to the cloud. Contact centers run in the cloud for a lot of real-time voice. And so things that we looked at before and said, those will never move to the cloud, have now moved to the cloud. And I think my third takeaway is just AWS is changing and they're now getting into areas to allow customers to do things they couldn't do before. So if you look at what they're doing in the area of AI, a lot of their AI and ML services before were prediction. And I'm not saying you needed AI, ML to do a prediction was certainly a lot more accurate, but now they're getting the generative data, right? So being able to create data where data didn't exist before, and that's a whole new use case for them. So AWS, I think, is actually, for all the might and power they've had, is actually stepping up and becoming a much different company now. Yeah, I mean, I had wrote that post. I had a one-on-one day, you know, got used to the transcript with Adam Sileski. He kind of went down that kind of route of, you know, hey, you know, we going to change next gen. Oh, that's my word. AWS classic, my word. You have AWS classic, the old school cloud, which a bunch of Lego blocks, and you got this new next gen cloud with the ecosystems emerging. So clearly it's Amazon shifting. Dave, your breaking analysis kind of teed out the keynote. You went into the whole cost recovery. We heard Adam talk about macro at the beginning of his keynote. He talked about economic impact, sustainability, big macro issues, and then he went into data and spent most of the time on the keynote on data, right tools, integration, governance, insights. You're all over that. You had that almost, your breaking analysis almost matched the keynote. Thematically, macro, cost savings, right sizing with the cloud. And, you know, last night, I was taught this one, the marketplace people, we think that the marketplace might be the center where people start managing their costs better. This could have an impact on the ecosystem if they're not in the marketplace. So, again, so much is going on. What's your analysis? Yeah, there's so much to unpack. A couple things. One is we get so much insight from the Cube community, plus your sit-down one-on-one with Adam Salipsky allowed us to sort of gather some nuggets and really, I think, predict pretty accurately. But the number one question I get, if I could sort of hit the escape key a bit, is what's going to be different in the Adam Salipsky era that was different from the Jassy era? Jassy was all about the primitives, right? The best cloud. And Salipsky's got to double down on that. So he's got to keep that going. Plus, he's got to do that end integration. And he's got to do the deeper business integration, sort of up the stack, if you will. And so when you're thinking about the keynote and the spirit of keynote analysis, we definitely heard, hey, more primitives, you know, more database features, more Graviton, you know, the network stuff, the HPC for, you know, Graviton for HPC. So, okay, check on that. We heard some better end to end integration between, you know, the elimination of ETL between Aurora and Redshift, Zs and I were sitting next to each other going, okay, it's about time. Okay, finally, we got that. So that's good, check. And then they called this thing the Amazon data zones, which was basically extending Redshift data sharing within your organization. So you can now do that. Now, I don't know if it works across regions. Well, they mentioned APIs and they have a data zone. And so I don't know if it works across regions. But the interesting thing there is he specifically mentioned integration with Snowflake and Tableau. And so that gets me to your point. At the end of the day, in order for Amazon, and this is why they win, to succeed, they've got to have this ecosystem really cranking. And that's something that is just the secret sauce of the business model. Yeah, and it's an integration into that ecosystem. I think it's an interesting trend that I've seen for customers where everybody wanted best to breed, everybody wanted disaggregated, and their customers are having trouble now putting those building blocks together. And then nobody created more building blocks than AWS. And so I think under Adam, what we're seeing is much more concerted effort to make it easier for customers to consume those building blocks in an easy way. And the AWXX, I talked to yesterday, they all committed to that. It's easy, easy, easy. And I think that's why. There's no question they've had. They've had a lead in Cloud for a long time, but if they're going to keep that, that needs to be up front. Well, you're close to this. How easy is it? But we're going to have Adrian Cockroft at the end of the day today go into him on analysis. Now he's retired. Well, that's difficult. How's that? Adrian retired from Amazon. He's a cube analyst, kind of retiree. But he had a good point. You can buy the bag of Lego blocks if you want, primitives, or you can buy the toy that's glued together. And it works, but it breaks. And you can't really manage it and you buy a new one. So, you know, his metaphor was, okay, the primitives allow you to construct durable solutions a lot harder relative to rolling your own, not like that. But also the simplest out of the box capability is what people want. They want solutions. Like we call that, I'm the solution CEO. So, I think you're going to start to see this purpose built specialized services allow the ecosystem to build those toys so that the customers can have an out of the box experience while having the option for the AWS Classic, which is if you want durability, you want to tune it, you want to manage it, that's the way to go for the hardcore. Not to be foundational, but I just see the solutions things being much very much like an out of the box. Okay, throw it away, buy a new toy. More and more and more, I'm seeing less customers want to be that hardcore, similar building blocks and obviously the really big companies do, but that line is moving. And more companies, I think just want to run their business and they want those pre-built solutions. We had to cut out of the keynote early, but I didn't hear a lot about, you know, the example that they often use is Amazon Connect, you know, the call center solution. I didn't hear a lot of that in the keynote. Maybe it's happening right now. But look, at the end of the day, sweets always win, you know, the best of breed does well, it takes off, you generate a couple billion, you know, Snowflake will grow, they'll get to 10 billion, but you look at Oracle, the sweets work, you know. What I found interesting about the keynote is that he had this dramatic exploration themes. First one was space that was like connected to Atheneblia, different kinds of the lens. Ocean. Ask the right questions. Ocean was security, which was more, you know, a lot more needed to manage that. Oxygen going deep, are you snorkeling, are you scuba diving? Barely interesting, right? And Antarctica, right? And Antarctica was the performance around how you handle tough conditions and you've got to get that performance. So we're laughing, but it was good. No, but the data ocean, Dave, I tweeted you, Dave, because you know, I sit on the cube in 2011. I hate data lakes, the worst term ever. It's the data oceans, more dynamic, there's a lot more flowing, maybe 10 years too soon, Dave. But, you know, he announces the ocean theme and then says we have a security lake. So like lake ocean, little fun on words. I actually think the security lake is pretty meaningful because, you know, we were listening to talk, coming over here, talking about it where, I think if you look at a lot of the existing solutions, security solutions, I describe them as a collection of data ponds that you can view through one map, but they're not really connected. And the amount of data that AWS holds now, arguably more than any other company, right? If they're not going to provide the security lake, who is? Well, but staying on security for a second. To me, the big difference between Azure and Amazon is the ecosystem. So CrowdStrike, Octa, Zscaler, name it, CyberArk, Rapid7, they're all part of this ecosystem whereas Microsoft competes with all of those guys. You know, so it's a lot more white space than the Amazon ecosystem. Well, I want to get you guys to take on something in your reaction because I think my vision of what's happening here is that I think that whole data portion is going to be data as code and I think the ecosystem harvests the data play. If you look at AWS' key announcements here, security lake, price performance, they're going to optimize for those kinds of services. Look at security. Okay, security lake, guard duty, EKS. I mean, that's a docker. Docker has security problems. They're going inside the container and looking at threat detection inside containers with Kubernetes as the runtime. I mean, that's a little nuanced point, but that's pretty significant, Dave. They're now getting into, we're talking in the weeds on the security piece, adding that to their large scale security footprint. Security is going to be one of those things where if you're not on the inside of their security play, you're probably going to be on the outside. And of course, the price performance is going to be the killer. The networking piece surprised me. They're continuing to innovate on the network. What does that mean for Cisco? I mean, so many questions. We had Ajay Patel on yesterday for you, and he's an awesome middleware guy. And I was asking about serverless and architectures. And he said, look, basically, serverless is great for stateless, but if you want to run state, you got to have control over the runtime. But the point he made was that people used to think of running containers with straight VMs versus Fargate or Knative if you choose, or serverless, they used to think of those as different architectures. And his point was they're all coming together. And it's kind of now you're architecting and calling which service you need. And that's how people are thinking about future architectures, which I think makes a lot of sense. If you're running managed Kubernetes, which everyone's doing, because no one's really building it in-house themselves, no, they're running it as managed service, skills gaps and variety of other reasons. This EKS protection is very interesting. They're managing inside and outside the container, which means that gives them visibility on both sides of the under the hood and inside the application layer. So very nuanced point, Z, is what's your reaction to this? And obviously the networking piece I'd love to get your thoughts on. Well, security obviously is becoming a, it's less about signatures and more about analytics, right? And so things happen inside the container and outside the container. And so their ability to look on both sides of that allows you to happen threats in time, but then also predict threats that could happen when you spin the container up. And the difficulty with the containers is they are ephemeral, right? It's not like a VM where it's a persistent workload that you can do analysis on. You need to know what's going on with the container almost before it spins up, right? And that's a much different task. So I do think the amount of work they're doing with the containers gives them that entry into that. And I think it's a good offering for them. On the network side, you know, they provide a lot of basic connectivity. I do think there's a role still for the Cisco's and the Aristos and companies like that to provide a layer of enhanced network services that connects multi-cloud because AWS is never going to do that. But they've certainly, they are as legitimate a network vendor as there is today. We had met up on yesterday, they were talking about latency in their product. I'll tell you this. In the analyst session, you know, Steven Armstrong said, you are going to hear us talk about multi-cloud. Yes. We're not going to mess their lead with it. But you said it before, never say never with Amazon, right? I mean, we talk about super cloud. You're like, Dave, ultimately the cloud guys are going to get into super cloud. They have to. They will do multi-cloud. I predict that they will do multi-cloud. I'll tell you why. Just like in networking. Well, customers are asking for it. Well, one, they have the, not by design, but by default, they're in multiple clouds are in their environment. They got to deal with that. I think that the super cloud and sky cloud vision is right. There will be common services. Remember networking back in the old days when Cisco broken as a startup. No one, there was no real shortest path first kind of thinking policy came in after you connected all the routers together. So, you know, right now it's going to be best to breed low latency, high performance. But I think it's going to be a need in the future saying, hey, I want to run my compute on the slower, lower cost compute. They already got segmentation by their announcements today. So I think you're going to see policy based kind of AI coming in where developers can look at common services across clouds and say, I want to lock in an SLA on latency and compute services. It won't be super fast compared to say on AWS with the next Graviton 10 or whatever comes out. So I think you're going to start to see that come in. Actually, I'm glad you brought Graviton up too because the work they're doing in Silicon, actually I think is, because I think the one thing AWS now understands some things are best optimized in Silicon Summit, software layers, something cloud and they're doing work in all those layers and Graviton to me is a home run. Yeah. I mean, Dave, they've got more instances. There's going to be, I mean, they already have Graviton's that slower than the other versions. So what are they going to sunset them? They paid, they don't deprecate anything ever. So, you know, Amazon paid $350 million. You know, people believe that it's a number for Annapurna, which is like one of the best acquisitions in history. And it's given them, let's put them on an arm curve for Silicon that is blowing away Intel. I mean, Intel's finally going to get, you know, Sapphire Rapids out in January. Meanwhile, you know, Amazon just keeps spinning out new Gravitons and Trainiums. And so they are on a price performance curve. And you know, like you say, no developer ever wants to run on slower hardware ever. Today, but if there's a common need for multi-cloud, they might say, hey, I got to trade off latency and performance on common services if that's what gets me there. There's maybe a business case to do that. Well, that's what they're... Which by the way, I want to, Salipski had a strong quote, I thought was if you're looking to tighten your belt, the cloud is the place to do it. I thought that was very strong. Yeah. And I think he's right. And then the other point I want to make on that is, I think, I don't know, I don't have any data on this, but I believe just based on, you know, some of the discussions I've had that most of Amazon's revenue is on demand, right? Okay, it's paid by the drink. Those on demand customers are at risk, because they can go somewhere else. So they're trying to get you into optimized pricing, whether it's reserved instances or one year or three year subscriptions. And so they're working really hard at doing that. My prediction on that is, that's a great point you brought up. My prediction is that the cost belt tightening is going to come from the, in the marketplace is going to be a major factor as companies want to get their belts tightened. How are they going to do that, Dave? They're going to go in the marketplace saying, hey, I already overpaid a three year commitment. Can I get some cohesity in there? Can I get some of this or that and the other thing? You're going to start to see the vendors and the ecosystem, if they're not in the marketplace, that's where I think the customers will go. There are other choices to either cut their supply or base or renegotiate. I think it's going to happen in the marketplace. Let's watch. I think we're going to watch that grow. I actually think the optimization services that AWS has to help customers lower spend is a secret sauce for them, that they, customers tell me all the time, that AWS comes in, they help bring the cost down and they want to spend more with them. And the other cloud providers don't do that and that has been almost a silver bullet for them to get customers to stay with them. And this is always the way. You drop the price of storage, you drop the price of memory, you drop the price of compute, people buy more. And then the question long-term is okay, and does AWS get commoditized? Is that where they're going? Or do they continue to thrive up the stack? John, you're always asking people about the bumper sticker. Hold on, before we get the bumper sticker, I want to get into what we missed, what they missed on the keynote. Yeah, there are some blind spots. I think, let's go around the horn and think, what did they miss? I'll start, I think they missed the developer productivity angle. Supply chain software was not talked about at all. We see that at all the other conferences. I thought that could have been weaved in. You mean security in the supply chain? Overall developer productivity has been one of the most constant themes I've seen in events. Who are building the apps? Who are the builders? What are they actually doing? Maybe Vernon will bring that up on his last day, but I don't hear Adam talk about it all, developer productivity. Would you, would you share? Yeah, I think on the security side, while they announced security data lake, I think the other cloud providers do a better job of providing insights on how they do security. With AWS, it's almost a black hole, and I know there's a careful line they walk between what they do, what their partners do, but I do think they could be a little clearer on how they operate, much like Azure and GCP. They announce a lot of stuff on how their operations works. I mean, I think platform across cloud is definitely a blind spot for these guys. I mean, I think you look at what you're doing. But none of the cloud providers have embraced that, right? It's true. I mean, maybe Google a little bit, Microsoft a little bit, certainly AWS hasn't at this point in time, but I think they perceive the likes of Mongo and Snowflake and Databricks and others as ISVs, and they're not. They're platform players that are building across clouds. They're leveraging, they're building super clouds. So I think that's an opportunity for the ecosystem, and I'm very curious to see how Amazon plays there down the stream. So, John, what do you think is the bumper sticker? You know, we're only in day one and a half here. What do you think so far the bumper sticker is for re-invent 2022? Well, to me, the day one is about infrastructure performance with the whole, what's in the data center, what's at the chip level. Today it was about data, specialized services, and security. I think that was the key theme here, and then that's going to sequence into how they're going to reorganize their ecosystem. They have a new leader, Rupa Bronis, who's going to be leading the charge. They've integrated all their bespoke, fragmented, partner network pieces into one leadership. That's going to be really important to hear that, and then finally, Werner, for developers, and, you know, event-based services, microservices, what that world's going on, because that's where the developers are, and ultimately they build the app. So you got infrastructure, data, specialized services, and security, machine learning with SWAMI is going to be huge, and again, how developers code it all up is going to be key. And is it the bag of Legos or the glued toy? So what do you want, out of the box, or do you want to build your own? And, you know, that's the bottom line, is that connecting those dots, all they've got to be is good enough, you know, I think, to your point, you know, if they're just good enough, less complicated, that will keep people on the base. I think the bumper stickers, the more you buy, the more you save, because from an operational perspective, they are trying to bring down the complexity level, and with their optimization services, and the way their credit model works, I do think they're trending now that back. My bumper sticker is Ecosystem, Ecosystem, Ecosystem. This company has 100,000 partners, and that is a business model secret weapon. All right, there it is. The keynote announced more analysis coming up. We're going to have the leader of AdMembers on this product, Marguerite Bibbo, coming up next, here to break down their perspective. You got theCUBE's analyst perspective here. Thanks for watching day two, more live coverage for the next two more days stay with us. I'm John Furrier with Dave Vellante and Zia's Cervella here on theCUBE, be right back.