This episode of Over the Edge features an interview between Matt Trifiro and Vishwamitra Nandlall, VP Technology Strategy & Ecosystems at Dell Technologies. Vish is an experienced CTO and a highly regarded telecom visionary. He is responsible for defining Dell’s technology strategy in the Big 6 domains, including 5G, Edge, Data Management, Cloud, AI and Security. In this episode, Vish discusses the future and excitement over augmented and virtual reality, and provides his thoughts on the future of edge computing.
This part 2 of 2 episodes of Over the Edge features an interview between Matt Trifiro and Vishwamitra Nandlall, VP Technology Strategy & Ecosystems at Dell Technologies. Vish is an experienced CTO and a highly regarded telecom visionary. He is responsible for defining Dell’s technology strategy in the Big 6 domains, including 5G, Edge, Data Management, Cloud, AI and Security. Widely recognized for his contributions to the industry, Vish has also held CTO executive leadership roles in Telecommunications for 25 years, including Telstra, Ericsson, Extreme and Nortel. Vishwamitra has been awarded a fundamental patent for LTE, published several widely cited technology papers, and holds several patents for the design of cloud based mobile applications and communication services.
In this episode, Vish discusses the future of technology and his excitement over augmented reality and virtual reality. He also talks about the future of edge computing, including everything from cost to security, and the realities of what’s needed for its advancement and further adoption in the tech world.
---------
“Real productivity comes at the intersection of people, labor and the types of tasks that they're trying to do. A lot of what edge can do is to start to mediate that intersection.”
“The real gift that edge is going to be able to deliver, is the realization of all these technologies that have had limited reach, are suddenly going to be able to spill out of the cyber domain and into the meet space.”
---------
(02:00) Open Grid Alliance
(04:15) Future of Technology
(13:00) Competition and Commercial Partnerships
(19:00) Considering the Customer
(21:30) Important Future Technological Focus
--------
Over the Edge is brought to you by Dell Technologies to unlock the potential of your infrastructure with edge solutions. From hardware and software to data and operations, across your entire multi-cloud environment, we’re here to help you simplify your edge so you can generate more value. Learn more by visiting DellTechnologies.com/SimplifyYourEdge for more information or click on the link in the show notes.
--------
Connect with Vishwamitra on LinkedIn
[00:00:00] Narrator 1: This part 2 of 2 episodes of Over the Edge features an interview between Matt Trifiro and Vishwamitra Nandlall, VP Technology Strategy & Ecosystems at Dell Technologies. Vish is an experienced CTO and a highly regarded telecom visionary. He is responsible for defining Dell’s technology strategy in the Big 6 domains, including 5G, Edge, Data Management, Cloud, AI and Security. Widely recognized for his contributions to the industry, Vish has also held CTO executive leadership roles in Telecommunications for 25 years, including Telstra, Ericsson, Extreme and Nortel. Vishwamitra has been awarded a fundamental patent for LTE, published several widely cited technology papers, and holds several patents for the design of cloud based mobile applications and communication services.
In this episode, Vish discusses the future of technology and his excitement over augmented reality and virtual reality. He also talks about the future of edge computing, including everything from cost to security, and the realities of what’s needed for its advancement and further adoption in the tech world.
[00:01:30] Narrator 2: Over the Edge is brought to you by Dell Technologies to unlock the potential of your infrastructure with edge solutions. From hardware and software to data and operations, across your entire multi-cloud environment, we’re here to help you simplify your edge so you can generate more value. Learn more by visiting Dell.com for more information or click on the link in the show notes.
[00:01:51] Narrator 1: And now, please enjoy this interview between Matt Trifiro and Vishwamitra Nandlall, VP Technology Strategy & Ecosystems at Dell Technologies.
[00:01:57] Matt Trifiro: Now Dell. Is one of the founding organizations of the open grid Alliance. In fact, this you, or one of the founding board members, tell us a little bit about why this concept of a grid or an open grid is important in fact to you, what, what is it and, and why is it important?
[00:02:15] Vishwamitra Nandlall: Earlier in our discussion, we were talking about this evolution of the internet. Yeah. And we we've moved towards this internet where we've centralized compute that creates pooling efficiencies, but it comes at the expense of propagation delay, but largely what fueled the internet was the growth of a digital economy.
That was self-contained right. It didn't have a lot of 10 rolls out into the real world. And I think what edge does it starts to put. A series of sensors into the real world and at bridges, this real world was this, this a cyber world. So that fusion, you know, you hear it come into different contexts, right?
Whether it's digital twin or, or metaverse, these are all cyber physical
[00:03:00] Matt Trifiro: systems or
[00:03:01] Vishwamitra Nandlall: systems that they all, they all relate largely to the same end goal of how do I bring the physical world and, you know, apply the, the means of productions of the digital world on that so that we can grow. Productivity overall for everyone.
So I think that that's kind of the origin of where edge comes in. It's largely in [00:03:20] order to, to meet the timescale realities of the physical world, I need to be closer to physical things that I'm governing. Yeah. In fact,
[00:03:26] Matt Trifiro: one of the best definitions of, of edge computing I've heard and I've heard them all and I've, I've made up some of them, myself it was from the, the CEO's Zen layer.
He said the edge is where. Digital meets the real world. And I think that's right. And the, the closer and physically, but also just, just temporally and from a, a business alignment, like closer in every sense is, is sort of the, the, the objective of edge, right? That like the internet just becomes the device and the device just becomes the internet.
There you go. I mean,
[00:04:02] Vishwamitra Nandlall: That to me is magic in, in some ways, right?
[00:04:05] Matt Trifiro: Yeah. So let's, let's talk, let's get the magic in two dimensions. So one is what this allows us to do. So let's focus on that first and then let's, let's back into what needs to be in place and built and collectively built in order to make that possible.
So what, what are some of the, the amazing things that you see coming into our, our future? If we do. In
[00:04:25] Vishwamitra Nandlall: terms of use cases, there's a lot of poster child use cases. The one that I'm very excited about is obviously AR and VR. And we know that there's, this is why,
[00:04:37] Matt Trifiro: why, why are you so excited about that? What, what
[00:04:39] Vishwamitra Nandlall: whats you, this, this opportunity to, to kinda live in a virtual world and interact with people in a virtual sense, kind of helps to build.
Helps to bridge the distance between different communities that are disparate parts of the world. The birth of the internet was called the death of distance, but in a real human sense, [00:05:00] that distance still still dominates another,
[00:05:02] Matt Trifiro: another shade of the meaning closer. It's interesting. It is. It's like socially and culturally closer.
[00:05:08] Vishwamitra Nandlall: Exactly. You know, to be able to bring an experience to someone's own, you know, geographic bubble. To me that that's, you know, akin to the vision of star Trek and hollow deck and all of these things where, you know, it's inspired generations of engineers. We finally have an opportunity to people to deliver that kind of experience.
So can we grow empathy on the back of that? I, I'm a big believer that meeting people builds empathy and building empathy, bridges differences, and, you know, not. put a picture of we're all holding hands and singing kumbaya, but I do think these things help, ultimately problems problems can be bridged through these types of these types of services.
So that to me is, is really interesting. I do think that I've long been very worried about the way technology has been developed. Over the course of, let's say even past two decades in the world economic form, we'll throw this out. And that we haven't really seen the productivity gains and is why there's definitely diffusion issues.
There's, you know, maybe it's unevenly
[00:06:10] Matt Trifiro: for every time saving macro. I've got five more emails in my inbox.
[00:06:13] Vishwamitra Nandlall: That's right. I think the reason why that's what happens. I think the reason why is that real productivity comes at the intersection of people labor. And the types of tasks that they're trying to do, a lot of what edge can do is to start to mediate that intersection.
So we might be able to drive more productivity gains. So if I think of cobots, which are collaborative robots that work alongside humans on factory floors, [00:06:40] that that ability to put in automation and that automation generates new tasks to increase and take advantage of the. Kind of the extra revenue that's captured as, as a result of that productivity and invested into new labor who, who can do new tasks and do high order tasks.
That becomes really interesting to me. And I think that that's something, again, from a use case perspective, when we look at smart factories and smart manufacturing, something that edge does help to deliver. And I think we're gonna see a lot of early entrance into the edge space from an industrial perspective, kind of be able to, to, to reap the dividends of that.
And I do think that it'll start. Ultimately track in terms of productivity gains, much more direct. Than what we've seen before, because we're now entering into the real world where services are done and services are rendered. And when it touches human labor and productivity, I think that's where we'll see the impact to GDP.
So I think that that, that could be the real gift that ed is gonna be able to deliver is the, the real realization. Of all these technologies that have had limited reach are suddenly gonna be able to spill out of the cyber domain and into the, into the meat space, if you will, and, and start to resolve some of that, that disconnect.
Yeah.
[00:07:55] Matt Trifiro: And I mean, I I've been in the, the edge space. I mean, For the last six or seven years in this new sort of wave of, of edge. And I co-founded the state of the edge. And so am right. I either can have credit or blame for establishing some of the canonical definitions around it. But one of the things that I've realized is that, is that the edge is more of an aspiration.
Well, it's, it's a, it's almost [00:08:20] a marketing. It's like, right. And it means the, all these things that you're talking about, which is how do we improve productivity? How do we make it less expensive? How do we make it more transparent, more integrated, more collaborative, more all these things. And it's not just.
The edge that's required. You, you require a lot of other things that, that may not exist literally at the edge, or they may need to exist at the edge, or they may be sort of orthogonal to the edge because it's like software orchestration or something, which is why a group of us went out and formed the open grid Alliance, which will say, okay, edge is important.
Cloud's important Internet's in between there's all these applications we want to build and deploy with this, these new technologies that are closer to the end point. How do we do that? Yeah. So what's, what's, what's your view of that?
[00:09:08] Vishwamitra Nandlall: You hit a, you hit a super important point, which is technologies.
Internet of themselves are a diamond dozen. You can call it edge or IOT or, or. It's really when you have the coevolution of that primary technology with a set of complimentary technologies that it takes root and is able to start to deliver benefits. The
[00:09:32] Matt Trifiro: internet of things doesn't do much use if the internet you're thinking is not real time and deterministic.
exactly. So
[00:09:37] Vishwamitra Nandlall: all, all these other things have to come into play. Yeah. You know, to your. you know what, so what are the enabling capabilities that we have to deliver? You know, again, I'll, I'll go back to some of the, the early insights I've gotten off of edge are this, one of them is what kind of workloads are best going to be optimized for an edge.[00:10:00]
And we know outta the box that there is going to be some things that are shaped by the economics of edge and then the utility of it. And those two things have. Kind of form an optimal point. And the things economically that we'll know about the edge is that the compute is just gonna be more expensive when it's distributed versus pooled.
That's just an economic reality. So the, the compute resources will be more expensive, but at the same time, we know that the transmission resources should be less. Cause I'm not going through as much. In order to deliver the content to an end user. And so the application probably doesn't need to be compute intensive, but it's probably bandwidth intensive.
One of the big things that, that, that satisfies is probably content distribution, which is why obviously content delivery networks were, were the first, very first edge. Yeah. Example. Yeah. But you know, you could think of. Live TV as being another enabler that you could see cloud gaming, which has long been pursued, but hasn't been realized Google stadia seems to be fading back.
And in fact, the, the players that seem to be doing well are not actual cloud players. It's companies who actually had the console like a end video GE force. So. I think what, what was lacking was enough density at the edge to be able to guarantee your assure the experience, because you do need a certain level of interactivity.
That's less than, let's say a hundred milliseconds and closer to maybe 20 milliseconds. So I think those two things are gonna be important in terms of understanding what's gonna be needed for the edges. What is the class of applications? The other side of [00:11:40] it is going to be, how can we make them more secure?
There's a physical security and then there's a cyber security issue around edge. And I think that that really starts to dominate in a lot of enterprise minds in terms of how am I gonna manage, you know, incorporating this into my it real estate. And that's an area that I think we're starting to get burgeoning understanding of.
We've we've obviously we've obviously got. The, the principles of zero trust that are helping us to, to govern that we, we are starting to look at things like distributed identity mechanisms in terms of concealing privacy, those types of things. I think that there's mechanisms here that we're, we're starting to, we're starting to shed light on and, and, and are starting to get implemented that are gonna be important for us to be able to really get the power of edge.
I think the other side of it is how do I get enough? Not just density of endpoints, but how do I get enough connectivity between edges typically that would be through some kind of equivalent, an IP X or equivalent of appearing point or exchange point for edge. We need to figure out how that's, how that's gonna scale and how that information's gonna exchange.
And how do we do that in a way that doesn't break the benefits of edge, where it's still driven largely by propagation delay. So, you know, we're definitely gonna have to look at different ways. Of tuning the networks. So that clusters of edges that have a relative latency to an end user that is similar, can be grouped and can form kind of a, a [00:13:20] distributed pool of service creation points that can deliver the experience to whether it's a machine or a human.
And I think that that's super important, how we architect. Oddly enough, we've done things like that before. If you look at like in any cast network that's being used in CDNs, these things are incredibly complicated to deliver, but they use similar concepts. You know, it's not clear to me that, you know, we're gonna be able to use anything outta the box to do this.
I think what we're gonna find. Is that there's a set of practices that will have to evolve over time and we need to share those operating practices to be able to build and tune these networks so that when I load balance across a cluster of equidistant edges, from a given point of use, that, that I can balance that and I can calculate that really, really quickly.
And then I can partition my network up, like. So that makes it scalable and it makes it deliver on the, the guarantees that, that we're, we're hoping to provide. So I think there's a bunch of technologies in there that, that need to be delivered at scale that we didn't necessarily have at scale with CDNs that were much more proprietary, siloed instances.
This is gonna be tough. This is not easy. I think that that itself is gonna be very, very tough.
[00:14:37] Matt Trifiro: One of, one of the things that I like most about the vision of the open grid Alliance is it's kind of coming at it. You talked about the slow way and the fast way, and, and it does seem like you do need both of these to push and pull on each other.
But in terms of like things happening at the edge, the slow way has been winning . So, so I love that the open grid Alliance is about, look, let's, let's go make [00:15:00] stuff work from a technical standpoint. A, but also a business standpoint and the, the point you made, just the complexity. I mean, we, in the innovation zone in Las Vegas, we stood up a very, very simple video inferencing over GCBs system on the grid in Las Vegas.
And there were no fewer than. 20 parties involved from the manufacturer of the radio, which like was the first one in north America for 5g P light Powell, spectrum CVS to the software company. That's, that's running the video management system and, and the Inre and everybody in between, right? From the Dell servers, you Amazon outpost to all of these things.
Right. And two things that struck me. One is just that, like the number of partners that. Converged to make something happen, but also that it did happen. This happened in six weeks that this came together. Yeah. Which is also inspiring. And so what I, what I see maybe as a, an end result of this, and I'd be interested in your opinion.
As you think about like, let's say factory automation or, or something like that. It's like today you go to one company and you buy a pretty bespoke solution that all sits on prem, go to Rockwell international or some system integrator. And you, and you, you get this thing and the advantages you've won neck to choke, as they say, and a, a bunch of work that goes into solving your particular use problem, but you don't get any of the advantages of.
Cloud economics, which what I really want, what I really want, I think, is a factory owners. I want a bunch of general purpose, open interface, six degree robots. And then I want to download the software that works great for it. That's right for my application. And maybe that's a little too [00:16:40] dreamy because making a car is different than making pencils, but.
I see that as the trend in the history, one company come and assemble all these vendors and like making it to something that works and that'll happen for a while, but eventually there should be some, there should be a lot of drag and drop. Like I wanna use this radio and this server and this software, and more or less, you should be able to assemble it without a lot of painful gymnastics.
And, but I'm interested in how you see that evolving.
[00:17:09] Vishwamitra Nandlall: Yeah, I think. You're you're hitting it. I'm kind of searching for the right words for it because there's lack of, you know, for a full business model. I always say that there needs to be kind of four elements, but three of them that are important, you need to understand how are you creating value?
How are you delivering value, how you're capturing value. And the force is kind of, you need to understand your basis of competition so you can understand relative differentiation mm-hmm and all those things are. I. in an economy like the hyperscaler economy. Let's say mm-hmm . Those elements are all well understood.
Everybody knows how they're creating value. Everybody knows how they're capturing value, capturing value through offering these services and developers are coming to it. And you're, you're able to capture value through the utilization of those underlying services, the VMs that are deployed on it, and the services that you're offering on top, whether they're database services or other you're charging for those, you know, maybe the, the, the basis of competition is, is largely the number of people that are using it.
The number of customers that you have, and everybody's lying for the number of workloads that they can host on. They're they're hyperscaler clouds. They understood. They can execute that. The, [00:18:20] the issue to your point around edge is that most of those are not well understood. So how do different members create value in edge?
So, what does a Dell do to create value? What does, what does an a IO do to create value and what are all our relative positions? You know, how do we each deliver value? How do we each capture value? Because we don't wanna stack our rates up and make this untenable, right? Yeah.
[00:18:44] Matt Trifiro: everybody wants 80% margin.
Why is nobody buying it? That's right.
[00:18:47] Vishwamitra Nandlall: So all of those rules haven't been shaped yet. Yeah. And they need to be formed through point solutions where people can start to grow commercial partnerships and then understand what those relative values are. But as more and more point solutions grow, the, the practices that are developed in one point solution start to become apparent across point.
So. And then we see things go from okay. Practices to best practices. And that's where yeah. You get that
[00:19:21] Matt Trifiro: might emerge in that, that particular yeah. Cuz the interface are well defined or the business relations well defined and the competitive pricing is, you know, that's right. Kind of known and yeah. And the disruption opportunities are maybe a little more obvious.
Yeah. Interesting. And I
[00:19:34] Vishwamitra Nandlall: just think we're, we're on the technology evolution curve for that. And we're in a period, not of execution, but of D. And those two phases of any kind of technology diffusion is important to understand. Right. And so the expectations for what we're gonna do now should be very different than what the expectations are when we're actually executing, because we're actually looking to get enough deployment out there so that there's some critical mass.
[00:20:00] Yeah. And so it's about winds and getting as much density of, of, of edge node as we can. And then it, and it's really starting to tweak out, well, what are the things that can be deployed at scale that have to, where are these
[00:20:13] Matt Trifiro: efficiencies and how can we, yes. How can we, well, we're kinda like in
[00:20:17] Vishwamitra Nandlall: the, you know, in a startup cycle, it's when you get product market traction, we're not really sure yeah, we do.
We don't have product market fit yet. We've got a bunch of ideas and hypotheses, and I think some are working, but it's uneven, it's uneven across deployments and I think. We're probably on a natural evolution to hit that in another three to five years where, you know, the discovery phase is gonna exit and then we're really gonna see it scale.
And the caution I give to people who think, well, wait, is that that's not the way the industry works. You don't. See that the ball linearly it's going to evolve, you know, kind of exponentially. Right,
[00:20:58] Matt Trifiro: right. It, it just hates and GTA and GTA, and then it explodes. And the people that were there, gest dating are the ones that benefit from the explosion you'll get
[00:21:05] Vishwamitra Nandlall: the dividend.
Right. And so that's the thing that it's kg to. To kind of explain to people. That's why I was kind of hesitating a little bit. I was going well, how do you explain this in the right way? We could explain it to a technology evolution, but I find most people can sympathize with that view.
[00:21:19] Matt Trifiro: I, I think one of the, one of the most powerful trends that technologists tend to ignore it.
And I'm sort of throwing myself into that, that group, which is just the very simple question, which is how does the customer buy it? That's right. And what do they have to go through and how much pain is it? That's and if you think. What customers want is they want a solution to the problem. [00:21:40] With no financial or, or technical or security risk.
Right. okay. And, and what that probably means is a turnkey product. That's been proven in the market that I can consume as a service. Yeah. And those aspects, there's so much of things that exist in the physical world that are just starting, you know, and you can, can you say that like Uber is a car as a service, right?
Well, for a lot of young kids, it actually is like, they don't wanna own a car. Like yeah. I'm much rather to have the person who knows. Own and operate cars, own the car. Just like I'd much rather have the person who knows how to own and operate servers own the server or things like that. And so it's it's and you see Dell responding to, to demands like that with the APECS program?
I mean, in the old days it looks like a lease, but it isn't, it's a, it's like I'm, I'm usually I'm going to Dell and I'm. Getting a server based on a page, you go relationship, which really is how I wanna own it. In fact, I probably don't even wanna manage it eventually. And that's a solution that's somewhere between you and the cloud provider.
That's that's there are different solutions for that.
[00:22:40] Vishwamitra Nandlall: Thank you. Got it. Yeah. I think when you're looking, I like the way you described it because it's yeah. How, how is a customer gonna wanna consume it? And they just, they, they want to. that's kind of the fuel versus friction discussion, right? They want all the features, but they want at no friction, they don't wanna have, they don't wanna have any issues in terms of, I gotta put, I gotta put a local it administrator on this and I gotta hire a bunch of people for $70,000 a year.
[00:23:03] Matt Trifiro: Um, yeah, let's hide all the complexity in the network and, and let the, the Dells and the Amazons and the Microsoft and the vapors of the world. Worry about it and let the enterprise just say, I want, I want building security. The. Yeah, I want
[00:23:16] Vishwamitra Nandlall: that. And it's done. And I think you're right. It's weird in terms of how these cycles work, [00:23:20] but usually the industry starts off as like an apple industry where everything's vertically integrated and it works.
And then it ends up as those things work, they start to modularize in different functions. Yeah. And that's where you scale. And then you get scale on the back end where it becomes the Android of the world. So it's a, it's a funny. It's a funny way to kind of navigate because you navigate with the end customer mind in the beginning, and you try to vertically integrate and put everything as turnkey.
And that means manage services and things that are probably a little bit more bespoke. But as you get to that inflection point where you need the economies to scale, things have to modularize and standardize. and we're kind of doing both at the same time, a little bit. Right. Okay.
[00:24:03] Matt Trifiro: My, my last question V.
This has been a great, great conversation. My last question. So if you look out into, into the near future, let's say the next, I don't know, 18 to 24 months. Yeah. And you had an opportunity to implore your. Colleagues, the other technology and other in, in companies that, that should they wanna take advantage of the inflection should be doing things now, what would you have them do?
Like what's the, what's the call to action to accelerate all of this that we talked about?
[00:24:33] Vishwamitra Nandlall: The, the thing that I think we we've done well is to put some of the fundamental stack in place for edge. And I think I'm actually pretty pleased with the way it's evolved. Mm-hmm I think what hasn't evolved. and maybe there's only one or two companies I can think of that have been focused on.
It comes down to what we were talking about earlier. How do you remove the friction? Its how do I abstract a lot of the distributed nature of edge? Hmm. Because that's [00:25:00] gonna be the, the thing that really kills it is the, the fact that you've got a sense, make the edge in a way that someone can visualize it.
And you've gotta much like how. I'm thinking of some of the work that's being done in w three C today was web GPU and it came out of WebGL and it may even web RTC those things when they come out. They're interesting, but they're only accessible by a limited pool of developers. And so the way they work is there's a bunch of abstractions and run times that are put on top of it, where it opens up the aperture of accessibility to broader pool of developers.
And the, the, the things that we haven't looked at for edge is how do we do, how do we do that? How do we make it? Like, if I think of how many developers there are in the world, the, the numbers kind of vary from 25 to 30 million. And so all the applications in the world have to be mediated through this group.
Yeah. And to the extent that they don't have accessibility to be able to develop applications and deploy applications to these edges is going to be a problem. Yeah. They're the, I see that. Yeah, they're fuel of the experiences. And so that's kind of what gets me a little bit worried is that these edges are highly heterogeneous.
These edges are highly complex in terms of the resources that are optimized for the different things, whether it's an AI accelerator, whether it's kind of a low power, processor's running unique kernel, whether it's, there's a lot of different things that. You know, an unevenness which require
[00:26:33] Matt Trifiro: and it's, and it's way too complex to reason about in a human brain.
It's way too complex. You need abstraction and [00:26:40] automation.
[00:26:40] Vishwamitra Nandlall: Yes. So those two things I think are important. One of the things, like I said, I, which I thought was, was, you know, a friend of mine, Jason Hoffman had a company called mobile inject. I know Jason really well, you know, Jason, and that was recently bought by Google.
But you know, that to me was one of those. Things that help to find and, and hone the resources and create
[00:26:59] Matt Trifiro: the Federation. They're trying to federate the edge and provide the uniform interface to that. Yes, that's right. Tapping the developer
[00:27:05] Vishwamitra Nandlall: community. I think that's a, that's an, a beginning piece, but there's so much more that we need to do.
Yeah. And it's really, how do I, how do I create that latter system for all these developers, to be able to create amplification, having a developer create the same application for a different edge. Is, I mean, we know how difficult that is just between smartphone ecosystems, let alone edge a class of edges that could burn the thousands.
So getting them to rewrite something a thousand times, that's not gonna work. So we need, we need some, some kind of abstraction layer. Some kind of run time that kind of enables that portability and obscures all that complexity, the, the power of the edge in some ways, counterintuitively is gonna be joint optimization of the hardware and the software, but for this get
[00:27:55] Matt Trifiro: without, without the software developer having to know that that happened.
yeah. So with
[00:27:58] Vishwamitra Nandlall: something we're never gonna get scale, if that's what we do. Yeah. But we have to figure out what's that middleware that does that. And like I said, there's, there's examples of us having been successful in. In the browser industry. I, I named a few. There it's even in, um, the different abstraction layers that are used for GPUs today, getting the developer community, being able to write AI applications for GPUs, took someone to write that.
[00:28:20] And you know, we're gonna need to take that up to the level of a compute platform called edge, and then move that even out in terms of how do you manage a cluster? Of likewise platforms that may each individually be heterogeneous, but can serve. Yeah,
[00:28:36] Matt Trifiro: what's interesting is my first instinct on who is in the best position to build that are the cloud providers and maybe VMware and maybe a couple of other companies like that.
There are hack Corp and the few companies that maybe could build this thing, or at least there are obvious candidates for, I'm sure there's a bunch of startups that could do it. And I think that. We do ourselves. And I mean, collectively the world, a disservice, if it's anyone cloud provider that does it in a non-open sourced way.
That's right. So that's gonna be an interesting tension cuz we need all the cloud providers there, but we also need them to recognize that multi-cloud is the future and that they. They're gonna have to fill differentiation, you know, not by locking an API, but by some other method,
[00:29:20] Vishwamitra Nandlall: I, I couldn't agree with you more.
I mean, to, to some, to some extent, this comes down to a values and principles thing as well. For me, I mean, I was part of the free internet movement of the nineties. And I, when I say free internet, I didn't mean that we should all get internet for free. It was, we need to make sure. No,
[00:29:38] Matt Trifiro: there are no artificial barriers of, of yeah.
Of great
[00:29:41] Vishwamitra Nandlall: hindrance control the internet. We can't have that. And I think, you know, the edge could be the last bastion of the open internet. It's kind of like, it'll be the Dyson sphere that surrounds the internet and we're gonna feed off of, because once an edge is in there, you can Pierce the opaqueness of all these different [00:30:00] provider networks and just do peer tope communications between.
Different edges. And that becomes your tour network. Yeah. That becomes your privacy. That becomes even a, a middle layer that allows you to get the same kind of services the big providers use today for their application delivery. There's all sorts of interesting things that could happen once we do that.
And I think, you know, it is really up to us as a community to start building that and laying the foundation for it. So that, that's why I'm excited about the O. Yeah,
[00:30:29] Matt Trifiro: that's really well. And it's also instructive that Amazon's a co-founding Amazon web service, a co-founder of the OJ too. So certainly they are as representative of that industry looking to move in that direction.
Hey, this has been a terrific conversation. I really appreciate you coming on the show and looking forward to maybe doing this again in a
[00:30:45] Vishwamitra Nandlall: year. Yeah, Matt. Hey, thanks a lot. I really appreciate it. And it was a lot of fun that
[00:30:49] Narrator 2: does it for this episode of over the edge. If you're enjoying the show, please leave a rating in a review and tell a friend.
Over the edges made possible through the Jenna sponsorship of our partners at Dell technologies. Simplify your edge so you can generate more value. Learn more by visiting dell.com.