An Edge Case | Ganesh Sundaram at AlefEdge

DISCLAIMER: Below is an AI generated transcript. There could be a few typos but it should be at least 90% accurate. Watch video or listen to the podcast for the full experience!

Ganesh Sundaram 0:00
Edges are movement. I mean, that’s the way to look at it. Edge is not something the internet was the movement. So the edge is the next generation of the internet.

Alexander Ferguson 0:18
As the speed and reliability of the Internet has improved, it’s become integrated into the basic functionality of our computer systems, phones, and even dishwashers and cars. But relying on massive data centers in faraway locations can slow things down. Enter the edge internet, a global framework that brings the servers you rely on closer to the edge of where you are. But this framework requires software solutions. This is the problem Ganesh Sundaram of Alefedge is trying to solve. In this episode of UpTech Report, I talk with Ganesh about his efforts to create an operating system for the edge Internet, and how he believes this technology will be affecting us all in the future. Good ash, I’m excited to be with you and learn more about LF edge. To start us off. I’m going to ask you to describe your company in five seconds. What is it?

Ganesh Sundaram 1:07
We are a software company, we focused on a category of the Internet called Software Defined mobile edge. And we started this company before it was cool.

Alexander Ferguson 1:20
This concept, maybe it’s been around. But it the focus wasn’t until really recently tell me how this evolution has happened and how you got to where you are today.

Ganesh Sundaram 1:32
From the company perspective mark, given my background working with in the mobility space, I used to be at Bell Labs in my previous life, and had an opportunity to participate and contribute towards the inventions that made 3g, 4g and all these technologies possible. So I saw an opportunity to be able to help carriers deal with the next generation of internet, which is how can they take advantage of the next generation of internet on contemporary access networks without having to upgrade? That’s kind of how it all started. And that is like is there a way to optimize the network rather than optimize the application? And another way of looking at it is, as the access networks and the connectivity networks have evolved from 3g to 4g to 5g, we were able to demonstrate that edge is a an integral part of 5g, as we all know, so 5g needs edge. But it doesn’t need 5g.

Alexander Ferguson 2:33
What problem are you seeing that you’re coming in to solve with? lf edge,

Ganesh Sundaram 2:37
the need the why part of the edge? Internet has been understood. Okay. But realizing the Internet, what is it that we need to do? And how are we going to achieve it is where software becomes a disruptive opportunity. So what is the software that is required to create a parallel connected universe, which I call as the edge internet that coexists with the existing internet and connects where needed? Okay. And the the example that I’d like to give is, telephony networks existed in the 70s and 80s. And then in the 90s, when cellular telephony started taking off, they created a parallel universe called the public land mobile network and connected with existing telephony network and tandem gateways. Okay, so the edges got kind of what should I say property where you can innovate using software. And so it is that software fabric layer around connectivity around cloud and application enablement, that is needed in this parallel universe, but at the same time connect with distributed data centers in the existing internet, so we may provide that software fabric to make connectivity meet computing.

Alexander Ferguson 4:03
So there’s this underlying hardware that’s in existence, but it’s the concept of the software on top in order to enable this edge opportunity edge computing that is kind of in tandem connected to 5g, your solution and you’re targeting what those who are managing the hardware, the mobility providers are always

Ganesh Sundaram 4:24
very good after. So who’s our customer? That’s a very good point. So it depends on the type of product line we’re going after. But generically speaking, our customers are going to be our enterprises who can take advantage of private edge. 5g, for example, one of the definitions of 5g is it’s a giant collection of private networks, which are all connected with each other. And so, enterprises want to control their own destiny. They don’t want to be dependent on operators, for example, carriers to be able to satisfy their connectivity needs. And they’re also very sensitive to the capex and optics, for example, a pharmaceutical company, their core competencies is is, is coming up with the next, you know, COVID, for example, vaccine for growth. So, their focus is not on optimizing network connectivity and making computing work at the edge. But there’s a there is a need, though, right? So, enterprises is one area of focus. But we can’t build all the software components. So we are also opening it up as API, so that a developer ecosystem can start to take advantage and contribute towards the edge software, fabric. So developers are also focused, on example, the third one is,

Alexander Ferguson 5:50
give me an example of the developers. What’s What’s something that could be developed that we could hook in using the API.

Ganesh Sundaram 5:57
Very simple. Let’s start with when in the connectivity layer, there’s going to be a collection of heterogeneous networks. Wi Fi is going to be there, there’s a new type of spectral access band called CB RS that has come up, there could be an existing 4g system that are emerging 5g systems, etc. And at the connectivity layer, we found a way to create a soft fabric. But there are legacy issues that already exist. And each one of them have different types of legacy issues. Can we create an API layer so that the usage metrics, you know what goes wrong, maybe a connective the ops layer can be built by by a smart developer out there who has the domain expertise on that. So that this is the go to market advantage we are creating for the customer. Edge is a movement. That, I mean, that’s the way to look at it. Edge is not something the internet was a movement. So the edge is the next generation of the internet. So there was, again, just to be cliche, it was a desktop internet, then there was the mobile internet, and now it’s the edge internet. That’s how I think about it.

Alexander Ferguson 7:15
So what do you see and kind of the next steps that would would simplify it or make or make it even easier for any enterprise or any organization say, okay, I can take advantage of this next wave of computing.

Ganesh Sundaram 7:31
Okay, so the cloud has given us a template, right? So go back 15 years ago, and close your eyes and start thinking about when enterprises were adopting your applications, right, they’d put out an RFP, 17, people will come back and there’ll be an RFP defense. And then by the time the decision is made, that technology is obsolete. The cloud made it completely seamless, like a whole bunch of API’s, the IT managers basically, try and then buy. Right? It’s all an API economy right now. So we’ve come up with this whole concept called as an edge API. Okay. So again, let me try to remind you of what how the cloud got adopted. The cloud was a very web based paradigm, then it was web was made programmable, and companies like Salesforce, and Amazon created what is known as a programmable Web. They build small software modules called Web Services, okay, and then expose those web services as web API’s. Right? So we have invented something called as the programmable edge, not just a programmable web now, can we push the barriers to a distributed internet architecture, and we’ve come up with something called as the programmable edge with edge services and API’s. So that’s one layer of friction that can be that can eliminate many layers of adoption friction, both on the technology dimension as well as the business dimension.

Alexander Ferguson 9:08
What have you seen the changes happen over this time? And and how many more years do you see until actually that grow even further growth and adoption happens?

Ganesh Sundaram 9:22
That’s a very important question. I can only revisit history from the from the, you know, a vantage point of how we have experienced it. Given that we started at the carrier edge by working with operator mobile operators. The carrier edge became a thing once 5g became a thing. So it was very rewarding to find out some of the early innovations that we had put in place, but actually adopted conceptually. In 5g, for example, on the architectural side is extremely rewarding for us. The The second aspect of the movement is the dimension around the Microsoft’s of the world, Microsoft, Amazon, any one of these hyper cloud companies moving towards a more distributed cloud architecture, which they call it the edge phenomenon. So Microsoft has something called as your edge zones. Amazon has outposts, local hub, sidewalk Greengrass, I mean, we just have a lot of investments that they’ve made on the edge. Okay. And then there are companies like packet, who we partner with, in fact, I mentioned my bunch of these names, because they’re all our partners in the packet, decided they were going to have the next generation of, of technologies for how to provide infrastructure as a service from the hardware perspective at the edge in a completely distributed manner. Okay. So I consider them to have done a lot of work, which is significantly important, and the market awarded them and Equinix acquired them for sensible sum of money, right? So, so the data center movement also took advantage of it. And then the third dimension of it is the new breed of applications has been constantly coming up, right? So there’s been a movement towards, for example, let’s take gaming, gaming has evolved from you know, having an Xbox in our, in our living room, and we are playing a game to the gaming controller is now in the cloud. So latency have become a major aspect of the story. Okay. So the concept of beamlets was invested invented by cloud by Google, for example. So the market dimension of the application, and then autonomous is another phenomenon, right? Autonomous cars, we’re not there yet. But the movement has created a new set of technologies in the application dimension. So it is the confluence of all these three dimensions, the connectivity, the application, and the compute dimension coming together. That basically demanded a more distributed architecture. I think, I think that’s why it happened so rapidly in a span of three to four years in my viewpoint. But you had a second question all along, is it going to take before it becomes a more regular phenomenon? Right. And I’m saying it’s happening now, meaning, we’ve transitioned from just doing turnkey projects to offering products at scale. It’s now a land grab moment right now. And so I have no doubt in my mind, and Gartner had a very interesting bell curve around around this. And we’ve been talking to Gartner a lot about these things. Also, it turns out that the hype cycle is in the next three to five years, this is going to happen in a big way. Okay. Now, only time will tell what application is going to be the priority. You know, four months ago, if we had had this discussion, I would never have brought up COVID-19 or conversations right? Now, it’s completely consuming our thought process is probably going to for the next several months. So what new business models are going to get created. But the economic underpinnings of that ask, have actually started to take the positive side of it. We feel it and we’re in the middle of so many deals, that bandwidth is a concern for us at this point. from a company perspective, good problem to try and

Alexander Ferguson 14:04
be sure to check out part two of my conversation with Ganesh, in which he dives deeper into the challenges of starting a movement around technology is still in its infancy. And he tells us the revelation he had when he examined how latency affects Instagram.