 Live from Orlando, Florida, it's theCUBE, covering Microsoft Ignite, brought to you by Cohesity, and theCUBE's ecosystem partners. Welcome back everyone to day two of theCUBE's live coverage of Microsoft Ignite. We are coming at you from the Orange County Civic Center in Orlando, Florida. I'm your host, Rebecca Knight, along with my co-host Stu Miniman. We're joined by Patrick Moorhead. He is the founder and president and principal analyst at Moor Insights and Strategy. Thank you so much for coming back on theCUBE. You're an esteemed CUBE alum, so. Gosh, this is great. Can you introduce me on every show, please? I would be happy to, delighted. So Patrick, before the cameras were rolling, we were talking about how many, frankly, tech shows you go to a year, we said 40, 45. That's about right. Yeah, I'm on, I live in Austin, but I actually live on a bunch of planes. It's kind of like you do, right? Right, sure, sure, yeah. So this is your 10th time at an Ignite, or an Ignite-like show, it used to be called TechEd. So what are your first sort of quick takes on what this conference, what you're seeing, what you're hearing? Yeah, so Microsoft has three layers, like a three-layer cake to their events. You have developers, you have customers, and you have channel. And this is their customer event. So what might seem like rehash or maybe build or inspire is you have customers who haven't heard this content before. So it's really about getting them engaged and things like that. And what we've heard first and foremost is we had 45 Azure announcements, but I think the biggest news was about the open data initiative that, I mean, how often do you have the three CEOs up on stage of where most corporate data sits with Microsoft, SAP, and Adobe? So it was impressive. And that's probably the number one thing so far. Okay, so let's dissect that a little bit. What are your thoughts? I mean, we're sort of questioning, it's a big idea. Right. When will customers actually see the benefit and is there a benefit to customers? So when I look at these big corporate announcements, I'm thinking, okay, is this thing paper? Is this thing real? How far does it go? I think this is real. When I dug under the covers in some NDA things that I can't give details on, there was meat there for sure. But where this all starts is two things are going on here. First of all, to do machine learning correctly, you have to have a lot of data, right? Yesterday's big data is today's machine learning. You have to have it all together. Now, you can pull in disparate data sources into your enterprise and work on that data, but it takes a lot of cleansing. Most of the time in machine learning is getting the data ready to be worked on. What having data interoperability standards means you can bring it in, you don't have to cleanse it as much, and you can do real-time analytics and machine learning on it. So it's agreement that says, we're all going to come in. If it's customer data, it's going to look like this with different fields. You would think something like XML could do this, but this is bigger. And from a competitive standpoint, I have to ask the big question, where's Salesforce and where's Oracle, right? They're the two odd companies out. Yeah, really interesting. You mentioned that there were a lot of Azure announcements here, something like 45. I was reading Corey Sanders had a blog of just, you know, listen, listen, list, and it's typical of what we see in the cloud. You and I, we go to AWS, reinvent, and it's like, oh, let's talk about all the compute instances, all the cool news storage, all the things, there's cheering and everything for every micro and macro, the thing that happens there. But are there any things that jumped out at you? We had Jeffery Snover on the program yesterday. We talked about the data boxes, like the edge and the various versions of those. Those seem kind of interesting when we talk about data and movement, but anything in kind of the Azure space that kind of got your attention. Yeah, so aside from the data boxes, I was really excited about AutoML. So, three ways you can do ML. You can do everything from scratch. You can take an off-the-shelf API, and then you can do something in the middle, which says, kind of like the free bears, right in the middle. Google at GCP announced something like this and so did Azure. And essentially what this is is it auto-tags your data. It's smart enough to know that this is an image as opposed to you having to start at the very beginning and hand-code some data. And it's not automatic because the key, so a good example might be an audio machine learning algorithm where you might need it for an airplane, versus a car, versus the factory floor, versus a smartphone application. Those are all different environments and your algorithm's going to be different. But as an enterprise, you might not have the PhD on staff to be able to do that, but you can't live with the off-the-shelf API. There's another thing that kind of struck me, a little bit of dissonance I saw there. You've got a Microsoft Surface sitting in front of you. Microsoft has gotten into hardware in a lot of places. When they talked about their IoT piece, they were like, we're going to put things out on the edge. And then on the other stream, it's like, well, but they're open and it's APIs and developers and software and not only Adobe and SAP, but now with Red Hat, talking about all they're doing with Linux, how do you reconcile the, I've heard people from Microsoft, we want to completely vertically integrate the stack and that's not something that I hear from the Googles and Amazons of the world. It was, I thought we're kind of past that. No one company can do it all. And the other hand, they're very open and give you a choice. How do you look at those things? So this all stems with the slowdown of Moore's law for general CPU compute. So as Moore's law is slowing down, we need to throw different kind of accelerators at the same problem to keep innovation going up and to the right at an increasingly faster pace. So people have gone to GPUs and CPUs and almost every one of the big infrastructure players has done that, right? Whether it's Google, Apple, AWS, that they all have their own hardware. Part of it is to accelerate time to market. The other is to get a lock in. I'm still trying to figure out which one this is. Now, Microsoft is saying very clearly in Azure IoT Edge that you can send your data even if you have their hardware to AWS and NGCP. And I think enterprises are going to take a quick look and I've been doing this almost 30 years. The gray hair, I have gray hairs to show for it, but you just have to pick your lock in, right? Enterprise IT always gets locked in and the question is what you lock in on. If you go with Oracle and then build applications around it, you're locked into Oracle. If you go with a certain hardware OEM, you could be locked into a certain OEM with converged infrastructure. So I think it's just picking the poison. You're going to have some people who are very comfortable with going all Microsoft and you'll have some people who want to piece part it together and look to the future. We still have people who were brought up on mainframes and they don't want to be there. They want to have flexibility and fluidity. One of the things you were talking about with the slowdown of Moore's Law, Microsoft, and frankly every other technology giant is really trying to stay ahead of the innovation curve. Microsoft, 42 years old, a middle-aged company and really in the tech world, a really old company, is Microsoft effective at this? I mean, do you see that this is a creative and ingenuitive and innovative company? Microsoft is one of the only companies that has been able to turn the corner from being aged and experienced, I guess like us, and moving into the new zone and everybody's work has had to do that. As an analyst used to, I remember getting Gartner and IDC reports on paper, right? But now it's very different. We're up here on theCUBE, we're on Twitter, we're doing research reports, right? So everything is changing and Microsoft has had to change too. And five years ago, when Azure hadn't really taken off, they had a billion dollar write down on Surface hardware, bought Nokia, shut Nokia down. You were wondering, wait a second, what really is happening? But then Satya came in and to the company's credit, it has completely turned around. And I will state though, you know, there's difference between perception and reality. I think a lot of the things that Balmer had in place were absolutely the right things. I think Satya gets a lot of credit for it, but these things just didn't magically appear when Satya came in. So a lot of things they did were right and it was perceived to be new leadership and therefore they're looking good. Yeah, I love it because we had quite a few Microsoft people on the program and a lot of them, 10, 20 years with the company. And they said, it's still the vision we had, but one articulated really well, I said we're even more focused on the customer than ever and that gets me really excited. Yeah. I want to ask you, when people look at this show, such a broad ecosystem, so many different pieces, what will they be talking about later in the year? My initial take coming out of it is, I'm a little surprised that we're talking so much about things like Windows 2019 and the Office 365, Microsoft 365, Dynamic 365. Obviously it's Microsoft strength, it's where they've got the most customers, but are the operating still relevant in the future? I met with the program manager of Windows 29 servers last night, Erin, and she had said that they had 1,300 people, they had to turn away from the Windows 2019 server. And it was 4,000 people, right? And I flippantly said, oh my gosh, I didn't think Microsoft still did that, right? It's all as a service, but I was just kidding, of course, but I think that shows how long it takes for people to move, but I think what we'll be talking about in a year is, has Microsoft delivered on its IoT commitments in IoT Azure Central? How much of their business has moved to, I'll call it on-prem software in a box, to as a service, right? So, Dynamics 365, Office 365. And then finally, I think we're going to see the workflow and here's something that my head finally went ding on, is Microsoft's strategy to surround the data and then do workflow on it to supplant Oracle SAP applications around the data. That's what I think we'll be talking about in a year. Yeah, one other specific I wanted to see if you've got some data on, because it's something we wanted to understand Azure Stack. The press oligog on it for the last couple of years. I really haven't talked to, I've talked to the partners that are working and people like Intel, Lenovo and the like, that are doing it, I haven't talked to too many customers that have applied service providers, yes. But what are you hearing, what are you seeing it? Is Azure Stack a big deal or is it just one of the pieces in the multi-cloud data application strategy that Microsoft has? So, Azure Stack is a big deal and I think that it's getting, it was a slow boil, to be honest with you. I mean, the company changed hardware strategies. It was first an ODM model and then it went to an OEM model and a very narrow OEM model. The compute requirements to Azure Stack were too big to some people, so it's a slow boil, but I look at what has the competition done. Now, to be even a public cloud player, you have to have an on-prem capability. With Google, it's PKE on-prem. You have Greengrass and Amazon DB that's on-prem sitting on top of VMware. So, hybrid cloud, multi-cloud is a real thing. I just think it's getting a little bit slower start than everybody had thought. Great. Well, Patrick, thank you so much for your insights. These were terrific. It was great having you on the show. Thanks for having me. I'm Rebecca Knight for Stu Miniman. We will have more from theCUBE's live coverage of Microsoft Ignite in just a little bit.