 Connectivity is an end-to-end problem. And open source connectivity is broken down into three sub foundations, LF Connectivity, LF Edge, and LF Networking. And they all solve one single critical infrastructure problem is how do you connect people? How do you connect things? How do you connect globally? And how do you connect with open source? And so today I'm going to talk about a few very interesting announcements in the world of networking that we just announced at the open source summit. Hi, this is Yoho Sapinbhartiya. And we are here at Open Source Summit in Bilboa, Spain. And today we have with us once again, Erpit Joshipura, GM of Networking Edge IoT and Energy at Linux Foundation. Erpit, once again, it's great to have you on the show. Great to be here. Yeah. And of course, this is an exciting, of course, we can see behind us great venue, Linux Foundation. You folks always pick great venues. Though sometimes it's hard to reach there, but there are great venues and great turnout, great audience. Talk a bit about this is first day of the event. What kind of audience, what kind of turnout you saw here? First of all, I agree with you. It's a great place, great venue. And our events team just puts up a fantastic show. So what I can say, though, that doing my keynote from the stage and looking at the audience, almost everybody had cameras up when I was speaking. So that felt good, a full house, obviously. But because of the content and because of the people that come to the events, it's really very exciting, especially with this one-on-one engagement in person finally after a long three years. So you're happy to see cameras because they are covering all the networking cameras they're all about Edge, they're all about IoT. That's why you're like, hey, this is my audience, right? This is my audience, this is, and they're consuming my projects. Exactly. Now, let's talk a bit about some announcements that came out because you wear so many hats, you let so many projects there. So we made three major announcements today. The first announcement was on a project called Kamara. Kamara is a global project that exposes APIs. APIs stand for application programming interface. So basically you are trying to get access into the network, the telco network, but then develop app on apps on top so that you can monetize that. So that we announced that project as a directed fund or a funded project so that we can scale. So that was one. Deutsche Telecom led the project and they were here on stage with me to announce it. The second announcement was on a very interesting project which is called Silver, which was formed and launched as an LFEU project. It influences global, obviously, but it does include a lot of interaction with LF networking projects like Anuket and it includes requirements from Europe that needs to be customized for European agencies or EUs. So that project is a cloud native implementation of a telco stack that is built on reference models and reference architectures that LFN's Anuket provides. So that was, again, led by Orange and joined me on the stage for announcing. And then the third announcement was we expanded our partnership, we as an LF networking, expanded our partnership with Etsy. So if you remember five years ago, we did a video on harmonizing standards and open source. And if Etsy has a standard, we would implement it. If they don't have a standard, we would code it and then have a standard created on that. So we were able to expand the partnership to include some very innovative projects like Nefio and Kamara as an example. I want to just briefly talk about Kamara. I talked to Marcus earlier, but I heard a lot of repeatedly about kind of how to monetize, how to commercialize. And then we look at open source, it is seen as two different word. Commercial word is different than open, but that is not reality without commercialization. Open source will not succeed, it will not survive. If you look at these events, or if you look at this project, I mean, Linux Foundation actually paved the path for corporate players to become good open source citizens. And that is why sustainable, that is why you see so much growth of that because we need that model. So can you talk about the importance of that? And then once again, the Kamara project there. You're asking two separate questions. There's a monetization or a commercialization of open source project in general. And I agree with you completely. We have paid the way of making what I call the plumbing layer of software common, right? Non-differentiated software, it's all common. And then companies, vendors, members, system integrators on top, they differentiate and then they sell it as a product, they support it, charge money for it, and they make money out of it. And that's the cycle that we support. And I think it's a proven cycle now. The second question on monetizing through Kamara is a different kind of monetization. What it is, network is full of rich data, okay? Rich data, right? Where phones are, what information exists, and they are the trusted partners. That data is not sold, right? They're not, right? But if you have APIs that you can provide services based on that data without necessarily pushing it out, then you can write applications on top. I think of this as an app store for the network. Like it could be some charging functions or it could be you are in a crowded area and you need bandwidth on demand or low latency on demand, right? Things like that. And then you can upsell those services. So that's kind of what Kamara is intending to do. And it's a different kind of monetization than the previous one. Networking and energy is the highways. If the power goes down, it doesn't really, your word comes to, same as networking as well, you know, connectivity. And if you look at EVs or any other, you know, Genetic AI, chat, GPT, talk a bit about the role of networking, not of careers telcos in modern, because it's going to play even more critical role as we are getting connected. And then let's talk about the role that LF networking is playing there. So let's first address the LF networking and LF energy question. Both of these are critical infrastructures, okay? US government, EU, right? Everybody understands that it's critical infrastructure because if either energy grid or network connectivity goes down, businesses, countries, you know, right? So given that this is critical infrastructure, how do you secure it? How do you operate it? How do you protect it, right? Is a high priority, right? For both organizations. The other thing I can tell you is LF energy and the energy grid is following the footsteps of networking. So five years ago, proprietary infrastructure, proprietary solutions, right? Very regulated industry, very standards-based. Same thing like telecom. Telecom moved to an open-source world now, fully disaggregated, SDN, right? Everything is there. Energy is the same. So LF energy is paving the way to do exactly what telecom did. And more importantly, they're also collaborating with LF edge through common projects like Fledge, right? Where they can use that project and their use case for the energy grid, right? At the edge of the energy grid. So they are doing, so there's a lot of commonality in terms of lessons learned, architecture, software solutions, standards and the behavior, right? That we are sharing as best practices. That's why kind of I'm running kind of the umbrellas. That's one. The second question on AI and how it is important is I want to make it extremely clear that we have four layers in AI. There is a infrastructure layer, which is the network. And there's tremendous amount of data that comes out from the network, right? Just we talked about that. But then what you do with the data has two parts. One is the data governance. How do you share the data? And you need to share it in a public setting or you can share it in a private setting, right? Which I call domain specific AI. And then there is the AI infrastructure, the GPUs that you run the models on, machine learning models on. That has both a centralized generic thing, like chat GPT, central, general public domain. And then there's domain specific, whether it's the energy domain or the telecom domain, right? So we are, LF networking is focusing on the telecom specific domain AI, both data sharing and the models. And then on top, you have the use cases. And the use cases are extremely custom for the domains, right? Customer billing or customer support or federated learning. Or, you know, there's about 10 use cases that our governing board has identified. And our governing board, LF networking, 80% of the top 10 CAPEX spenders are in the community, right? So we as a community are deciding what the top 10 use cases are and that's what we're gonna focus on. And these are domain specific AI use cases, right? Doesn't, so it's not chat GPT, it is domain specific, right? So just separate domain specific versus chat GPT. Because chat GPT, what it has done is it has just raised awareness that AI is important, right? Now we have to apply it to the domains. We have been talking like, and interesting thing in this case is that you start talking in like six months, one year that becomes an old technology, old topic, 5G. And I'm not just talking 5G, I'm also talking 5G private networks now. So now soon we'll start talking about 6G. So when we look at the progress that is going on, and a lot of times these are not technologies for the sake of new name and culture, but they bring a lot of use cases which are not possible because of the limitations of existing technology. They don't replace Wi-Fi or cellular, they complement them. So talk about what you're seeing for the next stage of 5G. Good question. So the first thing is 5G is in full deployment now and it's gonna be there for the next five years, okay? A lot of it is open source enabled. And the biggest thing 5G brought was lower latency, IoT connectivity, and that allowed for new use cases like private networks, connected cars, you know, things like that, right? In 4G we connected phones, in 5G we connected things, okay? 6G, which is probably 2028 and after, although we're working on it right now, is focused primarily on AI and how we can bring in AI to that equation, okay? Local learning, local implementation, local recovery, et cetera, et cetera. So that's already happening. People are only looking at it from a standards perspective. And that is gonna still continue rapidly in terms of deployment, but does not, as you said, replace 5G. 5G is still in the process of being implemented, okay? When we look at all these technologies, I do remember the early days of Edge, you know, when you look at your cell phone that you can barely get anything done. Now you can get 5G when you're traveling, LTE. And then you talk about 6G, of course, of few years ago, the US governments also released some bandwidth to further democratize it. I was thinking that by this time, all our laptops, they will have built-in chips, so we just open the laptop. It doesn't, you don't have to worry about connecting to Wi-Fi or something. When you look at 6G, or when you look at the future of connectivity, cars, they have built-in cellular. So talk a bit about 6G, not just from the perspective of technology, but the usage, how deeply it will go into our lives. So I think irrespective of 6G, I think the way you always connect is, you know, cheapest, fastest, first, and everything else second, okay? Why do you have so much, so many different alternatives for storage, right? Large storages are cheap, but they're slow. Fast storage likes flash and all, very fast, but expensive, right? Networking and connectivity is no different, okay? Wi-Fi, because of how it's set up, and if you can, would be the first choice, because spectrum is expensive, right? You're using shared commodity, right? Or shared, premium shared resources. So that's always the case. Now, where we have, where we are going, is actually very good, because there is the 3GPP route, which is 5G, 6G, you know, the standard spectrum, which is allocated and every country does differently, but it doesn't matter. It's a good, you know, all the spectrum and contiguous spectrums are good. So that's one path, and that's global macro connectivity. We also launched an umbrella called LF Connectivity, right, which was through a help of Meta, but there are three projects there, more coming, but there's an interesting, the reason we launched LF Connectivity about six months ago is to provide alternate connectivity access in areas which are challenged for this exact reason. So 60 gigahertz, there's a project called TerraGraph that would connect fixed wireless access, remote rural access, or very dense urban access with buildings, right? And it's in deployment in many cities today, right? CBRS is the other one, right? Like spectrum available on that. There are unlicensed technologies. So there's different ways. If you don't count ethernet and fiber and all the other fixed technologies, right? So that's, you have a basket of connectivity technologies and the way you use it is the most premium shared would be the last one. The free accessible is the first one. And what we do is we make the handoff and the interoperability very easy. Arpit, thank you so much for taking time out today as usual and talk about inviting your topics. I love talking to you. So I am already looking forward to our next conversation, but I really appreciate your time today. And thank you for throwing in unscripted questions, all of them. Thank you very much.