 Welcome to the great SuperCloud Debate, a power panel of three top technology industry analysts. Maribel Lopez is here, she's the founder and principal analyst at Lopez Research, Keith Townsend, CEO and founder of the CTO advisor in Sanjeev Mohan is the principal at Sanjmo. SuperCloud is a term that we've used to describe the future of cloud architectures. The idea is that SuperClouds are built on top of hyperscalar CAPEX infrastructure. And the idea is it goes beyond multi-cloud. The premise being that multi-cloud is primarily a symptom of multi-vendor or M&A or both and results in more stovepipes. We're going to talk about that. SuperCloud's meant to connote a new architecture that leverages the underlying primitives of hyperscale clouds, but hides and abstracts that complexity of each of their respective clouds and adds new value on top of that with services and a continuous experience of similar or identical experience across more than one cloud. People may say, hey, that's multi-cloud. We're going to talk about that as well. So with that as brief background, I'd like to first welcome our panelists. Guys, thanks so much for coming on theCUBE. It's great to see you all again. Great to be here. Thank you to be here. So I'm going to start with Maribel, what I just described. What's your reaction to that? Is it just like multi-cloud is supposed to be? Is that really what multi-cloud is? Do you agree with the premise that multi-cloud has really been, you know, what like Chuck Whitton from Dell calls it has been multi-cloud by default. I call it a symptom of multi-vendor. What's your take on what this is? Oh, wow, Dave, another term. Here we go, right? More to define for people. But okay, the reality is I agree that it's time for something new, something evolved, right? Whether we call that SuperCloud or something else, I don't want to really debate the term, but we need to move beyond where we are today in multi-cloud and into if we want to call it cloud five, multi-cloud two, whatever we want to call it, I believe that we're at the next generation that we have to define what that next generation is. But if you think about it, we went from public to private to hybrid to multi and every time you have a discussion with somebody about cloud, you spend 10 minutes defining what you're talking about. So this doesn't seem any different to me. So let's just go with SuperCloud for the moment and see where we go. And if you're interested after everybody else makes their comments, I got a few thoughts about what SuperCloud might mean as well. Yeah, great. So I agree with you. And like I said in a recent post, call it multi-cloud two dot oh, but it's something different is happening. And Sanjeev, I know you're not a big fan of buzzwords either, but I wonder if you could weigh in on this topic. By the way, Sanjeev is at the MIT CDO IQ conference, a great conference in Boston. And so he's a public place. So we're gonna have him mute his line when he's not speaking. Please go ahead. Yeah, so, you know, I come from a pedigree of being an analyst of firms that love inventing new terms. I am not a big fan of inventing new terms. I feel that when we come up with a new term, I spend all my time standing on a stage trying to define what it is. It takes me away from trying to solve the problem. So I find these terms to be words of convenience, like for example, big data. You know, big data to me may not mean anything, but big data connotes some of this modern way of handling vast volumes of data that traditional systems could not handle. So from that point of view, I'm completely okay with SuperCloud, but just inventing a new term is what I have called in my previous sessions, tyranny of jargons, where we have just too many jargons and they resonate with the IT people. They do not resonate with the business people. Business people care about the problem. They don't care about what we in IT call them. Yeah, and I think this is a really important point that you make. And by the way, we're not trying to create a new industry category per se. We leave that to Gartner. That's why actually I like SuperCloud because nobody's going to use that. No vendor's going to use the term SuperCloud. It's just too buzzy. So, but it brings up the point about practitioners. And so Keith, I want to bring you in. So what we've talked about, and I'll just sort of share some thoughts on the problems that we see and get your practitioner view. Most companies use multiple clouds. We all kind of agree on that, I think. And largely these clouds operate in silos and they have their own development environment, their own operating environment, different APIs, different primitives. And the functionality of a particular cloud doesn't necessarily extend to other clouds. So the problem is that increases friction for customers, increases cost, increases security risk. And so there's this promise, Maribel, multi-cloud 2.0, that's going to solve that problem. So Keith, my question to you is, is that an accurate description of the problem that practitioners face today? What did I miss? And I wonder if you could elaborate. So I think we'll get into some of the detail later on why this is a problem, specifically around technologies. But if we think about it in the abstract, most customers have their hands full dealing with one cloud. Like we'll, you know, through M&A and such and you zoom in and you look at companies that have multiple clouds or multi-cloud from result of M&A, M&A, M&A activity, you'll see that most of that is in silos. So organizationally, the customer may have multiple clouds but sub-organ silos, they're generally a single silo in a single cloud. So as you think about being able to take advantage of tooling across the multi-cloud and what Dave, you guys are calling the super cloud, this becomes a serious problem, just a skill problem. It's too much capability across too many things that look completely different from another. Okay, so- Dave, can I pick up on that? Please, I'd love, I was going to just go to you, Maribel, please chime in here. Okay, so if we think about what we're talking about with super cloud and what Keith just mentioned, remember when we went to see TCP IP and the whole idea was like, how do we get computers to talk to each other in a more standardized way? How do we get data to move in a more standardized way? I think that the problem we have with multi-cloud right now is that we don't have that. So I think that's sort of a ground level of getting us to your super cloud premise is that, and you know, Google's tried it with Anthos, like every hyperscaler has tried their right one to run anywhere, but that abstraction layer, you talk about whatever you want to call it is super necessary and it's sort of the foundation. So if you really think about it, we spent like 15 years or so building out all the various components of cloud. And now's the time to take it so that cloud is actually more of an operating model versus a place. There's at least a base level of it is vendor neutral. And then to your point, the value that's gonna be built on top of that, you know, people have been trying to commoditize the basic infrastructure for a while now. And I think that's what you're seeing in your super cloud, multi-cloud, whatever you want to call it, the infrastructure is the infrastructure. And then what would have been traditionally that pass layer and above is where we're gonna start to see some real innovation. But we still haven't gotten to that point where you can do visibility, observability, manageability across that really complex cloud stack that we have. The reason I love that TCPIP example of Sanjeev, is because it changed the industry and it had an ecosystem effect. And Sanjeev, the example that I, first example that I used was Snowflake, a company that you're very familiar with that is sort of hiding all that complexity. And so we're not there yet, but please chime in on this topic. You gotta mute again, off mute. Building upon what Maribel said, to me, this sounds like a multi-cloud operating system where you need that kind of a common set of primitives and layers because if you go in the typical multi-cloud process, you've got multiple identities and you can't have that. How can you govern if I have multiple identities, I don't have observability, I don't know what's going on across my different stacks. So to me, super cloud is that, call it single pane of glass or one way through which I'm unifying my experience, my technology interfaces, my integration, and I as an end user don't even care which cloud I'm in. It makes no difference to me. It makes a difference to the vendor. The vendor may say this is coming from AWS and this is coming from GCP or Azure, but to the end user, it is a consistent experience with consistent ID and observability and governance. That to me makes it a big difference. And so one of Floyer's contribution conversation was in order to have a super cloud, you got to have a super pass and like, oh boy, people are going to love that. But the point being that that allows a consistent developer experience and to Meribill's earlier point about GCP, it explodes the ecosystem because the ecosystem can now write to that super pass if you will those APIs. So Keith, do you buy that number one and number two? Do you see that industries, financial services and healthcare are actually going to own clouds or what we call super clouds? So Stingy hit on a really key aspect of this is identity, let's make this real. Dave, you love talking about data collaboration and I love Stingy's point on the business user kind of doesn't care if this is AWS versus super cloud versus et cetera. I was collaborating with a client and he wanted to send a video file and the video file, his organization's access control policy didn't allow him to upload or share the file from their preferred platform. So he had to go out to another cloud provider and create yet another identity for that data on the cloud. Same data, different identity. A proper super cloud will enable me to simply saying as a end user, here's a set of data or data sets and I want to share collaboration, a collaborator and that requires cross identity or cross multiple clouds. So even before we get to the past layer and the APIs we have to solve the most basic problem, which is data. How do we stop data scientists from shipping snowballs to a location because we can't figure out the identity. We're duplicating the same data within the same cloud because we can't share identity across customer accounts or et cetera. We have to solve these basic thoughts before we get to super cloud. Otherwise we get to us a turtles all the way down thing. We'll get into snowflake and what snowflake can do but that's what happens when I want to share my snowflake data across multiple clouds to a different platform. Yeah, you have to go inside the snowflake cloud which is leading through, right? So I would say to Keith's questions, Sanjeev, snowflake I think is solving that problem but then he brings up the other problem which is what if I want to share data outside the snowflake cloud. So that gets to the point of, is it open? Is it closed? So Sanjeev chime in on the sort of snowflake example and then Maribel, I wonder if there are networking examples because that's, as Keith's saying, you got to fix the plumbing before you get these higher level abstractions but Sanjeev first. Yeah, so I actually want to go and talk a little bit about network but from a data analytics point of view. So I never built upon what Keith said. So I want to give an example. Let's say I am getting fantastic web logs and I know how much time they're spending on my web pages and which pages they're looking at. So I have all of that. Now all of that is going into cloud A. Now it turns out that I use Google Analytics or maybe I use Adobe's analytics suite. Now that is giving me the business view and I'm trying to do customer journey analytics and guess what? I now have two separate identities, two separate products, two separate clouds. If I, as an IT person, no problem. I can solve any problem by writing tons of code but why would I do that? If I can have that super pass or a multi-cloud layer where I've got a single way of looking at my network traffic, my customer metrics and I can do my customer journey analytics, it solves a huge problem. And then I can share that data with my partners so they can see data about their products which is a combination of data from different clouds. Great, thank you. Maribel, please. I think we're having a load of the rings moment here with the one ring to rule them all concept. And I'm not sure that anybody's actually insented to do that, right? So I think there's two levels of the stack. I think in the basic, we're talking a lot about, we don't have the basic fundamentals of how do you move data, authenticate data, secure data, do data lineage, all that stuff across different clouds, right? We haven't even spoken, right now I feel like we're really just talking about the public cloud venue and we haven't even pulled in the fact that people are doing hybrid cloud, right? So hybrid cloud, then you're talking about, you got hardware vendors and you got hyper scalar vendors and there's two or three different ways of doing things. So I honestly think that something will emerge. Like if we think about where we are in technology today, it's almost like we need back to that operating system that Sanjeev was talking about, like we need a next generation operating system. Like nobody wants to build the cloud mouse driver of the 21st century over and over again, right? We need something like that as a foundation layer. But then on top of it, there's obviously a lot of opportunity to build differentiation. Like when I think back on what happened with cloud, Amazon remained, AWS remained very powerful and popular because people invested in building things on Amazon, they created a platform and it took a while for anybody else to catch up to that or to have that kind of presence. And I still feel that way when I talk to companies. But having said that, I talked to a retail the other day and they were like, hey, we spent a long time building an abstraction layer on top of the clouds so that our developed works could basically right once and run anywhere. But they were a massive global presence retailer. That's not something that everybody can do. So I think that we are still missing a gap. I don't know if that exactly answers your question, but I do feel like we're kind of in this chicken and egg thing which comes first and nobody wants to necessarily invest in like, oh, well, Amazon has built a way to do this. So we're all just going to do it the Amazon way, right? It seems like that's not going to work either. But I think you bring up a really important point which there is going to be no one ring to rule them all. You're going to have VMware is going to solve its multi-cloud problems. Snowflake has a very specific purpose build system for itself. Databricks is going to do its thing and it's going to be more open source. Companies like Aviatric, I would say Cisco even is going to go out and solve this problem. Dell showed at Dell Tech World, a thing called Project Alpine, which is basically storage across clouds. There are going to be many super clouds. We're going to get maybe super cloud stove pipes. But the point is, however, for a specific problem and a set of use cases, they will be addressing those and solving incremental value. So Keith, maybe we won't have that single cloud operating system, but we'll have multiple ones. What are your thoughts on that? Yeah, we're definitely going to have multiple ones. There is no community large enough or influential enough to push a design, take Mirabelle's example of the mega retailer, they've solved it, but that's their competitive advantage. They're not going to share that with the rest of us and open source that and force that upon the industry via just agreement from everyone else. So we're not going to get the level of collaboration either originated by the cloud provider or originated from user groups that solves this problem big for us. We will get silos in which this problem is solved. We'll get groups working together inside of maybe a industry or subgroups within the industry to say that, hey, we're going to share or federate identity across our three or four or five or a dozen organizations to be able to share data. We're going to solve that data problem. But in the same individual organizations and another part of the super cloud problem are going to, again, just be silos. I can't run machine learning against my web assets for the community group that I run because that's not part of the working group that solved a different data science problem. So yes, we're going to have these bifurcations and forks within the super cloud. The question is, where is the focus for each individual organization? Where do I point my smart people in what problems they solve? Okay, I want to throw out a premise and get you guys reaction to it, because I think this, again, I go back to the Meribos TCP IP example, it changed the industry, it opened up an ecosystem. And to me, this is what digital transformation is all about. You've got now industry participants, Mark Andreessen says every company is a software company. You've now got industry participants. And here's some examples, it's not, I wouldn't call them true super clouds yet, but Walmart's doing their hybrid thing with Azure. You got Goldman Sachs announced at the last reinvent that it's going to take its tools, its software, its data, which is on-prem and connect that to the AWS cloud and actually deliver a service. Capital One, we saw Sanjeev at the Snowflake Summit is taking their tooling and doing it now, granted just within Snowflake and AWS, but I fully expect them to expand that across other clouds. These are industry examples, Capital One software is the name of the division that are now, to the reason why I don't get so worried that we're not solving the Lord of the Rings problem that Maribel mentioned is because it opens up tremendous opportunities for companies. We got like just under five minutes left. I want to throw that out there and see what you guys think. No, I was just, I want to build upon what Maribel said. I love what she said. You're not going to build a mouse driver. So if multi-cloud, super cloud is a multi-cloud OS, the mouse driver would be identity or maybe it's data quality. And to each point, that data quality is not going to come from a single vendor. That is going to come from a different vendor whose job is to harmonize data because there might be, data might be for the same identity, but maybe at different granularity level. So you cannot just mix and match. So you need to have some sort of like resolution and that is an example of a driver for multi-cloud. Interesting, okay. So, you know, octa might be the identity cloud or Z-scaler might be the security cloud or Calibra has its cloud, et cetera. Any thoughts on that Keith or Maribel? Yeah, so let's talk about where the practical challenges run into this. We did some really great research that was sponsored by one of the large cloud providers in which we took all, we looked at all the VMware cloud solutions. When I say VMware cloud, VMware has a lot of products across multi-cloud now in the product portfolio, but we're talking about the OG solution, VMware vSphere. It would seem like on paper, if I put VMware vSphere in each cloud, that is therefore a super cloud. I think we would all agree to that in principle. What we found in our research was that when we put hands on keyboard, the differences of the clouds show themselves in the training gap and that skills gap between the clouds show themselves. If I needed to expose, let's our favorite friend, a TCP IP address to the public internet, that is a different process on each one of the clouds that needs to be done on each one of the clouds and not abstracted in VMware vSphere. So as we look at the nuance, yes, we can give the big controls, but where the capital ones, the JP Morgan Chase just spent $2 billion on this type of capability. Where the spin effort is done is taking it from that 80% to that 90, 95% experience and that's where the effort and money is spent on. That last mile. Maribel, we're out of time, but please bring us home, give us your closing thoughts. Hey, I think we're still going to be working on what the multi-cloud thing is for a while. And super cloud I think is a direction of the future of cloud computing, but we got some real problems to solve around authentication, identity, data lineage, data security. So I think those are going to be sort of the tactical things that we're working on for the next couple of years. Right, guys, always a pleasure having you on theCUBE. I hope we see you around. Keith, I understand you're bringing your Airstream to VMworld or VMware Explorer. You put it on the floor. I can't wait to see that. And Mrs. CTO advisor, I'm sure it'll be by your side. So looking forward to that. Hopefully Sanjeev and Maribel, we'll see you on the circuit as well. Yes, definitely. I'll see you there. I can't forward to hopefully even doing some content with you guys at VMworld Explorer too. Awesome. All right, keep it right there for more content from super cloud 22. Right back.