 So our next two presenters and panelists, I would say, fireside chat, it's a different format. I would like to call on stage Marissa Viveros from IBM and Mohammad Awad from ARM on the stage, please. And I will let them introduce them. Please give them a round of applause. All right. Why don't you go ahead and introduce yourself? And? Great. So good morning, and I repeat in the whole next thing, great to be here, and thank you for having us in this important topic. So Marissa Viveros, I lead within IBM all of the strategy and solution development for the telco and media industry. This is a significant part of the business within IBM. And as part of my role, I've been leading within the past three years, all of the IBM play and strategy within network financial revitalization and now telco, network cloud. So this is a very appropriate team. In addition to that, I'm a member of the LFN board. And I just wanted to reinforce that IBM has a tremendous commitment from open source and throughout the presentation and throughout this minister we speak here, we will see how that has been represented within the market. Mohan? Thanks. Thanks, Arfit. Thanks, Marissa. So my name is Mohamed Awad. I manage the infrastructure business for ARM. We say infrastructure. We mean everything from the edge of the network all the way into the cloud data center and then into HPC. It's an area where we are putting a tremendous amount of focus and really excited about the changes happening in the ecosystem today. All right. And while we are on that topic, I have a few slides for you. Do you mind telling us your thought process and then we'll get into the discussion? Sure. OK. So ARM to date has shipped over 125 billion processors have shipped with ARM technology. Over the last five or six years, more than half that number was shipped. The point is that's accelerating. We're seeing more and more devices shipped every year. And those devices are more and more connectivity as a requirement. It's being driven a lot by IoT. It's being driven a lot by all these smart, intelligent end points looking to put data back on the network. That's changing the way the infrastructure works. We're moving from an infrastructure where, which has primarily been driven and designed around pushing data out from the core of a data center out to billions of people with phones in their hands, watching those cat videos on Facebook or YouTube or whatever. 70% of that data traffic is video today. But in a world of a trillion devices, which we believe will happen, that model changes significantly. And I want to just go through one quick example. If we talk about an HD video sensor, a billion HD video sensors, so in a world of a trillion devices, you could imagine that a billion of those devices are HD video sensors. Not unlike your Nest video camera at home. A billion HD video sensors would generate about 400 exabytes of monthly bandwidth. 400 exabytes of monthly bandwidth is about three times the total internet bandwidth today. Total internet bandwidth today is about 150 exabytes. So a billion HD video sensors would swamp the current infrastructure. We take over 40 million servers to manage all that data. That's more than the entirety of the AWS cloud. The point that I'm trying to make here is that the infrastructure is changing. As more and more of these devices get on the infrastructure, how we think about the infrastructure and the network has to change. Ultimately, what it means is that all along the path, whether it be that endpoint all the way up to the cloud and every point in between, data will have to be filtered, it will have to be analyzed, and it will have to be processed. This is changing the underlying compute platform all the way across. No longer can we rely on one centralized general purpose compute at the center of the data center to do all of that compute and routing and traffic management. It's going to have to happen everywhere. What open source does is it leads the way for that modularity and portability. It leads the way so that an application developer doesn't have to worry about where his application is running. All he needs to worry about is what is the latency, what is the bandwidth, and what is the compute. Is my service provider going to live up to those expectations. And that workload can then migrate across a vast array of different devices and platforms all throughout that infrastructure, whether it be at the endpoint, at the edge, or into the cloud. We've seen this happen already. We've seen people go off and make decisions where they go off and optimize different areas of that stack, whether it be open source software in the middle which enables vendors to go off and create customized hardware which can drive more and more acceleration at the lowest levels. We saw that with Amazon where they're doing their own custom cores now on custom silicon. But the application developer who sits on top doesn't matter to them. And that's really the point. The point is that you can get more customization, you can get more acceleration, you can get more efficiency by creating fit for purpose hardware underneath. Fit for purpose devices all the way throughout and workloads can move from one end to the other. We're seeing this happen more and more. And it really is the only way that we're going to be able to live up to the demands of this next generation, trillion device infrastructure. Thank you. All right, Marisa, just the same way. We thought it would be good to have some slides. I know a panel typically, if we just keep talking, then you'll be on the phone. So this is a better idea. Good. So let me just give you a little background. So in case for a lot of you that do not know, IBM works with about 84% of the largest providers across the world. That work, it is across helping them really become digital organizations. And when I mean digital, it is not digitizing the front end processes, the support processes, how they deal with clients. And part of that equation is the network. I would say the fundamental part of that equation is the network. Therefore, digitizing the network and making that network software, it becomes more important than ever. RP asked us to look at business models here. How does ecosystem business models are changing in this new environment? And I thought I start by saying we did a survey about six months ago. And we interview about 200 executives throughout the world. These are all telecom executives. Half of them are still able execs. So we asked them, how important is open source for them, open standards, open source in the revolution of their business, as well as the lifecycle automation of services that are being deployed? As you can see from the statistics from the results here, customizing it is very important. So having the flexibility and the agility to provide new services in that environment becomes critical. The second area is lower development costs. And that means really tap into the shared R&D that is happening with open source in order to get the economy of a scale, be more agile, and develop a new service in a faster manner. For me, one of the key findings here is faster time to market. I think every one of the operators, what they want is the faster time to market, that agility to bring new services, to test services in an agile manner, in an MVP, in a minimal viable product manner, test, look at whether it is successful, and then go back and rework if it needs refinement. But these are some of the characteristics of why operators are looking at and willing to adopt open source within interoperability, again becomes a critical element of that open source adoption. So these are the areas that I think as technologies and as the community here, we're all working on. What I like to shift a little bit are focus. And I say shift focus, and I really mean balance focus from infrastructure to service delivery. I think as a community, we need to pay attention to the network and how we build the network in a software environment. We need to pay attention to how we tap into the multi-cloud environment that everybody is looking at right now. How do we tap into machine learning, AI, and all of the new technologies that are available, not just for the sake of technology, but for the idea to provide that extremely good user experience to the end services. So that's what the operators want from us as a community, as companies, provide a great experience. Furthermore, bring that great user experience into verticals. No, the revenues that are being produced from consumers, we all know that those are declining on a daily basis throughout the world. Operators need to tap into, how do they tap into IoT, and other technologies, and Noesh, and 5G, in order to provide those new solutions for enterprises and really transform health care, retail, energy and utilities, you name it, all of the industries are there to be changed by this vast availability of bandwidth, lower latency within this network. So pay attention to not only the infrastructure, but also how we drive that from a service delivery perspective and what services we want to deploy. I think that's fascinating. The two things I really like is CAPEX is not in the top three, meaning coming from a vendor myself all my life. I really feel good that open source is not killing a business model for vendors. It's clear from your participation and your increased contribution. There's money to be made, and what the end users are after is not cheap. They're after innovation. They're after time to market, flexibility, customization. So that's a good. And the second thing I picked up, which was really good, was the service delivery or the service experience aspect. I think we heard Armagan yesterday who also stretched our imagination saying, that's kind of what the CXO people are looking at. And so for us and the networking deep dive land, for us, a user experience is basically, we get very excited when we move from CLI to a GUI. Let's be honest. It's like, oh, wow. We don't have to do CLI. That's GUI. That's user experience has gone up. Take that and stretch it 10 times. That's not what you're talking about. So with that, let me go through a few questions. The first one, Muhammad, we're in a world of heterogeneous compute. And that could be from the cloud to the devices. And how do you guide your company as well as the ecosystem to kind of balance, on one hand, its interoperability, compatibility, standardization. And the other hand is differentiation. So what are your thoughts on that? Yeah, I mean, at the end of the day, you can have heterogeneous compute, but it can still exist within a standardized framework. And I think that's really what is required and really what drives, in many cases, all new business models in and of itself. Being able to manage multiple different pieces of hardware, whether it's that ruggedized end node, which is in a power constrained environment, or that big HPC server, which is sitting in the core of the data center, being able to manage those two and then quickly deploy workloads from one end of the infrastructure to the other, becomes of itself a challenge. And we are seeing companies and products emerge to kind of address those things. I think that the key is, and one of the most important pieces of it, is having that standardized open source software in the middle, puts the onus on the developer of the particular pieces of hardware or the software to ensure that interoperability across the board. So in ARM, we contribute over 100 open source, just within the infrastructure business unit, over 100 open source standard projects. And we do that specifically to ensure that an ARM based server will work as seamlessly as a server based on some other architecture. OK. Yeah, so essentially what you're saying is, keep the plumbing level common. Exactly. I hate to use that word, but that's the reality. So the plumbing level is common, layer is common, that's based on open source. And then you add your customization and differentiation. That's right. And because it's open source, it can quickly be modified and adapted to ensure that it stays common. And it's no coincidence that we have the two ends of the stack on a panel. If you look at, and we are all networking folks, always stacks, right? He's at the bottom of the stack, right? That's what they always tell me. Bites and chips. And she's at the top of the stack. Services, things like that. So how do you see innovation? He would prefer things to be common at the hardware, API, abstraction, OS, device, driver, kind of that level. You would like to see it at the services layer, at the OSS-BSS layer. So how do you think innovation and interoperability can be balanced? I think the previous panel did a great job and therefore all communities do a great job at the standard. So while we do work at the upper layers of the stack, OSS-BSS, as well as the business services that are being delivered, we trust that there will be that common infrastructure that will allow our traders to innovate faster. So I think it is really that ecosystem that it really needs to work together. And for me, sometimes we are competitors. Sometimes we're partners. But that it continues to be more and more interlaced because the software needs to run on some hardware. No matter how much we dream, the software will never run on air. So the hardware will be there to stay. And we need to have it. But that having that common infrastructure at the lower cost that will enable in the ability to have new services at that cost. Accelerations and all those are needed. And can you give a few examples of some of the values that have been created in the ecosystem for the end users? So one of the key examples that we have been working with in IBM is one with Vodafone. And with Vodafone in Europe, we are doing several projects. One of them is we're working with them in the 5G Milan testbed. And there are about 100 companies that have come together to enable working with the Italian government and not really sponsored by the Italian government, working to create an environment where we have the infrastructure but new services can be delivered. So if you go to the website, you will see about 100 companies dealing from health care companies, insurance companies, all of the technologies, all of the service integrators are there, all of the network equipment providers are there, as well as the ship providers. So it's very, very amazing how these companies are all working together in bringing that unisum and accelerating the 5G deployment within this. Very nice. Very nice. Any examples from your perspective on either end user value created in a customer or in a project that comes to mind from Europe? I'm sure there's like hundreds of. Yeah, there's lots of them. I think one that I gave earlier was the AWS Gravitron announcement where they did, which in that case, what they did is because of all the work that was done on the open source side, they were basically able to go off and create their own silicon with custom acceleration in there, which then can go be transparent to the application developers, but can provide a 45% cost reduction as opposed to other architectures. One thing we have not talked about a lot is the barriers. And specifically, I want to point out that some of the larger operators have people. They have vendors that they can choke their necks, et cetera. What about the smaller end users? How do we service them? What can this community do to help drive that? Any ideas there? Maybe I'll start with you. Yes, so in my experience in working with operators throughout the last couple of years, specifically on network transformation, I see the larger operators. And even I think I can count with my fingers the operators that do have engineering staff, that they do have developers. And they can rely on to adopting open source testing and validating which ones are good for them. What I see is a huge gap between those and the rest of the world, where there is no engineering resources, where the engineers are so scarce that I think we as a community need to simplify what they should be adopting, so providing a component that they can easily adopt. But I think that is also in thinking about the topic is, I think if we can do also ourselves, we can be ambassadors to college universities and start bringing that team and bringing the knowledge into those new talents that we can harvest from and then create the next generation pipeline that can help bring to life those new networks. Very nice. And one last round, and I know we're going to run out of time, but on security, I know this is at the guts of the system that is extremely critical. And I know you're doing everything in your power, including collaboration with your competitors in the space to make sure that security is well-addressed. I know we will not have time to go through it, but give us a few sound bites on it. Yeah, I mean, I think the thing about security, it's an important topic, much more than I'm going to be able to cover in a couple minutes here. But it starts at the bit and bite level, transcends the device across architectures, acceleration, all the way up to the service level, and then all the way from the device, all the way back to that core, that data center. At ARM, that's the kind of the way that we think about it. We've done things like PSA and TrustZone, but we don't just think about it at the device level. We actively look at what does it mean from the device all the way back to that core, the data center, and how do we tie all that together so that we can create a uniform amount of security. OK, so a holistic view. And we have to take a holistic view of security. It's going to require a lot of standardization. It's going to require a lot of collaboration, both with us and then quite candidly with our competitors to make sure it's done right. Awesome. Mohamed, I will add security and trust, transparency is a key attribute that everybody is looking for. Well said. Well, very good with that. I think we're out of time, but really appreciate your insights and thank you for the support. Please give a big round of applause.