 Good afternoon, Cube fans, and welcome back to beautiful Barcelona, Spain, where we're here for four days of coverage at Mobile World Congress. It's midday through day two. My name's Savannah Peterson, joined by my fellow analysts, Dave Vellante and Shelly Kramer. Nice to have you both next to me. Very excited for our next conversation to kick off our post lunch content with someone who's been at the edge for 20 years. Sadaf, welcome to the show. Thanks for being here. Thank you very much. I'm so glad to be here. Versus on the edge. We're all on the edge of the stage right now, including this wonderful foliage. We're all on the edge. We want to hear about the Edge Compute Stack that you just launched. Absolutely, so I've been in this space for a while now, folks, so 20 years, 20 years plus in this space, I started and ran the IoT business at Cisco and then was running the Intelligent Edge business at Hewlett Packard Enterprise for several years. So at VMware, we have been looking at our customer's challenges in the space which is Edge, right? So it is a very exciting space because it's very close to the outcomes that the customers are chasing. But when one of our customers wants to go and deploy Compute at the Edge or digitize the Edge, it's not a very simple and easy path. It's a complicated path. So think about a retailer, right? So for example, if they want to reduce theft in their stores and they want to go in and deploy technology for that, they have to worry about a lot of things. The idea, going from the idea of theft reduction, going to actually realizing it, it's a long and difficult path. There you have to worry about what software you're going to use, how you're going to run that software, what hardware you're going to use. Even when you figure those things out, you have to worry about deploying. If you have 5,000 stores, how do you make it all happen in those 5,000 stores? And let's say that you are lucky and you were able to deploy, how do you now run it? That is a very challenging task. So my mission and goal and my vision for VMware by Broadcom is to become that company that solves this problem. So the Edge Compute Stack that you mentioned, Sherry, that Edge Compute Stack essentially solves for this problem. It makes it very easy for our customers to go from the idea to actual outcome in a seamless and pain-free manner. So instead of worrying for months, years, how I'm actually going to make all this happen through digitization and computer at the Edge, this platform brings all that together in a simple, easy way for our customers to deliver. So that's what we're doing. And Edge Compute Stack is a platform that we're using to make that happen. So Edge, Edge is not new, right? The Edge has always been there. We think of it as this physical place. But when I think Edge, I always think of CDNs. I think of companies like Akamai. But there seems to be a new Edge. So how should we be thinking about the evolution of Edge? No, absolutely. That's a great question, right? So the Edge, the evolution of Edge has been through several phases, right? So when I started off in this space, there was a lot of skepticism that was in the industry, whether you could actually digitalize the Edge, the plant floors, the retail stores, the hospitals, and could you actually put technology in these environments and get something out of it to realize their outcomes? So if it's patient outcomes, or whether it is the sentiment of the customers on a retail store, there was a lot of skepticism around that. We have come a long way since that time. People now understand that technology can be used, people can leverage it, and it can actually help them achieve their outcomes. So that time has evolved. We have gotten to the outcomes very effectively. What's happening now is that we are focused on experiences. So our customers are no longer coming to us and saying, hey, just give me an outcome. They're beyond that point. They know the outcome can be achieved. What they want is for that outcome to be achieved with an experience which is stellar, which is something that will hold people to that outcome for the long term. That's what we are delivering, and that's the evolution that we have been through over the last 20, 25 years in the Edge space. It sounds like your clients are more aware of their own personal user experience in that case. Diving in there a little bit farther, since it's a very hot topic here at the show. How is AI affecting this at the Edge? Yeah, so that's a great question. So AI and Edge are pretty much the two hottest things that are out there right now. Everybody wants to do AI, and interestingly, everybody also wants to do AI at the Edge. At the Edge, exactly. So I'll answer that question in two ways, right? One is the easy angle, right? So AI is absolutely going to help you do better in terms of deploying computer at the Edge, doing connectivity, right? So figuring out where computer is needed, moving it from box to box, and making sure that it's optimized for the applications that are running, making sure that the connectivity is also optimized by machine learning, right? Figuring out which connections work at which time best from a pricing perspective and all that. So that's the first piece, right? The second piece is more interesting. So our customers want to actually use artificial intelligence in these Edge environments. Now, the challenge with Edge is that this is not like your data center environment. Edge has constraints. There are space constraints, the devices are smaller, power connectivity is intermittent, sometimes it's there, sometimes it's not, sometimes it's very expensive, so on and so forth. There's these constraints you have to work with. So the question becomes that if you want to deploy AI in these environments, how do you make it happen? So for that, right, we have various models that we've been working with our customers on. One has to do with pruning of the models, right? So if the models are very big, it's going to be very difficult to deploy. Another thing that we've been working with our customers is a concept that we call the federated learning, right? So the way it works is that if you, let's say have 5,000 factories, 5,000 is a big number, 5,000 retail stores, for example. And you want to basically learn from all these locations and in a central place, and you want to create a model for let's say doing sentiment analysis of clients on the retail floor. There's going to be a lot of data that will have to be pumped up. You probably don't have the bandwidth for that. You probably don't have bandwidth all the time for that if you have stores in rural areas, for example. If you're a factory, you might have factories in areas that are far away from cities where there's no connectivity. So you'll have a problem. So the way federated learning works is that you learn in all these locations. Let's say you have 100 factories, you're learning in all these 100 factories separately. And then we have this product called VMware Edge Cloud Orchestrator, which is part of our Edge Compute Stack portfolio, which essentially brings all these learnings together into one super learning, learned model, right? So the trained model is one super model, which is a combination of all the learning that takes place in these edges. So that's how we are essentially helping our customers deploy AI at the edge in a way which works in these constrained environments. That's the future customers wanted. They want AI to happen very close to where the data is being produced, but it's been a problem for them doing it so far. With this Edge Compute Stack platform, you're solving that problem for them. That's awesome. Well, and speed. Speed has to be of the essence in this environment too. I mean, I love this retail example. I advise a retail theft company down in New Zealand, and it's such a big problem, and most people are repeat offenders. So having that data right away for the security guard in that retail environment as an example, to be able to do something about it before it's too late is imperative. How often does speed come up? We talk about inference a lot on the show. How is speed a huge priority for your clients? Speed is huge. Speed is everything in these environments. So I'll give you another example. Retail, that was an excellent example. I'll give you the example of manufacturing. In the manufacturing on the plant floor, we are now working with our customers. Audi is one of our big customers. We are virtualizing the programmable logic controllers, PLCs, which are at the heart of any plant floor. They run the operation, they move the robots, they do the welding, they do all those things. Now, what happens is that, let's say that you have a robot, which is swinging, and a camera detects is the person in its path. The robot needs to stop. And you cannot rely on connectivity outside the manufacturing plant for that decision to be made. You need for that robot to stop there and then. Instant. It's real time. It has to be real time. So what we are doing is that profanet, the protocol that these people run on these factory floor, we are implementing a real time version of that in the stack, in the edge compute stack. So you can have that safety feature available to your customers on the plant floor and you don't have to worry about connectivity, going out there here. Everything is going to be done instantaneously. So anytime I hear PLC. Yes. I think of Stuxnet. And so I have to ask you about security. Absolutely. At the edge. You know, there's a lot of companies that have startups that have been founded on the premise of edge security. Yes. Others are saying, no, you got to consolidate security. That's the right approach. Right. And they're both right. So how should we think about security at the edge? That's a great question. It is something. So I have a very heavy background in security. So I come from that area, risk management, how to improve asset availability and make sure that assets stay secure. So it's very close to my heart. I'll tell you that the way that we think about security and risk management at the edge is different in these operational environments from anything else that we do in IT. So in the IT space, we used to think of security in the terms of CIA, confidentiality, integrity, and availability in that order. This is not how things work in an operational environment. On a factory floor, A comes first. First is availability. Do not bring my factory floor down. Yes. Keep the factory floor running. Yes, absolutely. Integrity, yes, important. I don't want the wrong commands to be sent to the wrong machine. Confidentiality, yeah, sure, maybe. If we need it, we can have it. Ultimately, in these environments, delivering the outcome is paramount. So the way that we are thinking about in these environments about security starts from safety. So safety is paramount for them. They want to make sure that their employees are safe, their customers are safe. They also want to be compliant in terms of regulation. And so everything that we are doing is keeping these things in mind and then delivering to the outcome. So outcome is what is being secured here as compared to the assets that you have in place. So it's a different way of thinking. Outcome being secured slash preserved, I'm hearing. Absolutely, yeah, preserved. So speaking of life-saving security. Yes. I know that over at your booths, you have an ambulance. And I have a feeling that there's a story there about maybe AI, maybe about Edge, maybe about how we're using this technology to save lives, I don't know. Yeah, yeah. Is there a story? There is, there is Shelley, there is a story. So let me tell you a little bit about how we got into hospitals, right? So what the year is, how the IT industry has gotten into hospitals. So typically when you think of IoT and Edge, right, there are two kind of extremes. There is the enterprise edges, right? Like your office locations where knowledge workers work, right? So they take advantage of knowledge, information, and they do things with that. And then there's these operational environments such as oil and gas fees, plant floors, retail environments, which are very heavily operationalized. There's equipment, there's things happening. So those are two extreme environments that are very different from each other. Hospitals are interesting. They're kind of a mix. There are hybrid environments somewhere in the middle, right? There are knowledge workers, doctors, technicians who are kind of trying to figure out what to do in these environments using the information they have. And then there's a lot of equipment that goes into a hospital to make sure the patients are taken care of, right? So it's a combination of these two. So what we are doing a demo here is not just any hospital, but it's a hospital on wheels. So it's an ambulance that we have actually brought compute into. So that just like a hospital, we would bring IoT and compute into that hospital to enable the knowledge workers as well as the operational aspect of things. Same way in this ambulance, we have compute in there, which does a couple of things. From a knowledge worker perspective, we have virtual reality, right? So you can do a, you can wear VR goggles and the person, the technician in the ambulance who is perhaps trying to resuscitate somebody or trying to do some procedure can talk to an expert back in the hospital and that expert can see what is going on in the ambulance right there and then. Provide guidance, provide feedback and tell the technician what the right things might be to save the situation. Second thing that we are doing is more on the operational side, right? So there's equipment in the ambulance. Just imagine what would happen if some key pieces of that equipment were missing when you left to save somebody's life. Yeah. So the second piece is the operational. Oh, that just gave me chills to think about. Whoa, yeah. If you have the solution in the ambulance, that will make sure that all the pieces that are supposed to be there are actually there. So those, that's how we bring together the two aspects, the knowledge worker, the enterprise edge aspect and the operational edge aspect together in one place and deliver to our customers. So that's our hospital on wheels that you're seeing at MWC this year. Yeah, that's very, very cool. So tell us as we wrap here, what does the future look like for Edge? Where do you see this going? Yeah, so to me, so Edge is my career, it's my passion. So the coming years, I see a lot of improvement in terms of the experience that customers get. We have solved for the outcomes already and now we're very focused on delivering stellar experiences to our administrators, to our end user from Edge environment. So you're going to see that happening. The pain with deploying compute at the Edge, the pain associated with making that work in various types of environments, constrained environments. You're going to take that away and make it super easy for somebody trying to get to an outcome, to achieve it in a very quick time, get to the ROI without all the pain that is associated. Where is the developer at the Edge? One of the debates that we were having is, I know GSMA has this open API initiative, but maybe it's just because it's so big, but at AWS re-invent, you can't help but bump into a developer. Same thing with so many of these shows. KubeCon, same thing. What does the developer at the Edge look like? Who is that individual? So that's a great question. So let me tell you a little bit about what I'm seeing from a developer community perspective at the Edge. So we see customers in three states of transition at the Edge. They're customers who are doing some initial connectivity, some digitalization at the Edge. That's the first stage. The second stage is where they figured that out. Now they're trying to do something a little bit more fancy. They're trying to virtualize things. They're trying to connect things together. They're running multiple apps in a single environment. They're trying to do a little bit more complex environment. Then we have customers in the third stage who actually have active developer communities. They build a platform. They build a platform that runs on the Edge and they now have active developer communities who are building on top of that applications for their business. I'll give you an example of that. We have a very large customer of ours in UK. They're one of the largest retailers in the world. You'll be surprised to know that this retailer has 5,000 strong developer community within their company. This is a retailer. This is a grocery store. And what they have done is that they built a platform that allows them to gather customer sentiment data that allows them to use computer vision in their stores, to do theft reduction, to improve the experience of their people walking into these stores. And these developers are actually sitting there, writing apps, building apps, which allow all this to happen in an innovative way. And they're bringing in people from vendors such as us to help them make this developer community their job easier. Edge is very much about the developer community. Edge native apps that you hear about nowadays from cloud native to edge native. How do you take advantage of the distributed nature of the Edge, proximity to the data? All of that is being exploited by the developer community. The reason I ask is, Shelly and I and some other researchers at the Q Research have been working on this idea of an intelligent data app, intelligent data. In the old days of computing, everybody used to develop their own apps. And then this thing, they had the funny name called COTS, commercial off the shelf software came along and then SAS and okay. It seems like that example you gave is people are inventing new capabilities. And the idea is what struck me is earlier you said, people thought it was aspirational to actually digitize the Edge. And our research suggests that people want to have a digital representation of their entire business in real time, people, places and things. From the Edge all the way back to the core and the cloud in real time. And that has to be developed. That doesn't exist today. We call it Uber for the enterprise, right? And so is that the type of vision that you think can become a reality given all the constraints that you've talked about at the Edge? No, absolutely. I do believe that is the case. In fact, I'll tell you that as more focus comes to the Edge, people will start removing some of these constraints. The constraints that are associated with the size of the compute, the space availability, the power, right? They will see the benefits, that proximity to the data, the speed at which the decision making takes place. That's so paramount that they will realize that they need to get rid of these constraints that we have in the Edge environment. Not all of them can always be removed, right? There are some constraints that are physical. In, on a battlefield, there are going to be constraints that you'll probably never be able to get rid of. But many of these can be removed. So I'll, but I'll share a perspective with you, right? I'll go to manufacturing as a vertical, that's a key vertical where there's a lot of digitization starting to happen. In manufacturing, the concept of people developing their apps and deploying them on the factory floor and doing things, that used to be a very foreign concept just a few years back. There are customers now, very large customers. I don't want to name them right now because I don't want to put them on spot, but the very large customers who have products, vehicles, all over the place who now have people in their factory floors who are writing AI apps, various types of intelligent apps to do things in a different way in their factory environments. This was unthinkable. Just five to 10 years back. Because these environments are so controlled. Don't touch it. Don't touch it. Exactly. How do I touch it so it gets better? Yeah, that's good. It's a different way of thinking, but still, people want to make sure that they don't break what is intact, so we are helping them do it in a way that makes sense. Thank you. Oh, you seem like the perfect partner for that. I have one final question, and this is for Satat, the person, not the big brain in the corporate enterprise worlds. Given that you get to see so many different edge applications and solutions and ideas in their formative stage, what has you personally most excited about our edge future? I think for me, what has me excited, the most excited about the edge future, is the new generation coming in after us, right? So a lot of the folks who are coming into IT, they have been used to the environments where they're going into these big tech companies and building a future there. That's all good, right? But now they have another avenue. So with all the digitization that has happened at the edge in retail stores, in factories, in plants, in oil and gas fields, now they don't have to go the traditional path of going into a tech company. They can build a career for themselves closer to the actual business, closer to the actual outcomes. So if they're interested in an oil and gas environment, for example, and they're interested in that that fascinates them, they don't have to ditch technology to go there. Technology is there, they can bring all that together and they can be part of that transformation. So that excites me. Yeah, me too. I think we're ushering in a new era of entrepreneurs and to your point, we're going to be able to creatively fulfill our dreams in a different way that don't require some of the asylum paths that we had in the past. So you are an outstanding guest. Thank you so much for joining us. Thank you for having me. Shelly, Dave, always appreciate your insights and fabulous questions. And thank all of you at home or wherever you are around the world for tuning into our fabulous four days of coverage here at Mobile World Congress in Barcelona, Spain. My name's Savannah Peterson and you're watching theCUBE, the leader in enterprise tech coverage.