 I love Linux conferences because I usually get to say whatever I want and I will do that today. I'm going to talk about not what Intel does in open source and I'm not going to talk about open source either because and the reason for this is because I think open source is already one so kind of talking about open source is no longer needed and I know this came up few times and I think open source one because everybody realizes that hey we want to focus on our own business and let's build all the common stuff and infrastructure and operating systems and do that in open source where it's common everybody contributes to it. From an Intel perspective it's simple you know we want we want our customers to have the best experience running any software on our hardware and our customers want to run open source so that's why we contribute very happily to open source and specifically to Linux. Okay so today I'm going to talk about I'm going to talk about new projects and new ideas that we're starting to work on or we're barely you know or projects that are barely started so and this is more on taking an end-to-end view on edge computing and cloud-to-edge computing and where we do feel that open source is the primary software that will drive this revolution. So data you know when you hear about the billions and billions of devices those billions of devices the one thing that is common is that they will generate massive amount of data so when you think of a factory or a self-driving car is a perfect example self-driving car generates four terabytes a day this is just massive amount of data and that's great but but as you know that this data gets created it gets transmitted it needs to get stored it needs to get processed it needs to get acted on and and this is where you know just you know taking this massive amount of data back and forth to the cloud isn't really really practical so so this is where edge computing or computing on the edge and computing doing a lot of computing on the device becomes really really very important and that requires with the with what everybody expects that yes there is going to be 50 billion devices smart devices that are connected not just dumb toasters you know 200 billion connected sensors generating all this data you know there is certain amount of processing and certain amount of on device actions that need to be taken that drives different types of software that we in open source we need to figure out how to start working on so when you look at all of these things whether it's autonomous drones or cars or any of those type of things you can think a car is a self-driving car is a perfect example a self-driving car will have some hundred computers inside of that and it's effectively is a data center of its own it's a just driving data center or a flying data center and so on and and the type of software that you do need in order to fulfill that that that promise that that promises to radically also change the economy so I'll give you just one simple example in cars if if somebody can figure out the software update in cars several automakers did an actual estimate of like how much would it save us collectively if we figure out how to do over-the-air software update for cars for software in cars the number that came back was 35 billion dollars so there is tremendous there is like fundamental change in the economy and fundamental change in how people live their lives and the entire concept of what what safety is and so on and and these things require truly you know different types of software that we that we need to think about so few things that that we are looking at you know and and some of these may be the same as existing technologies and some of them are just like completely new containers you know same you look at orchestration well kind of functional safety we don't know really what to do we should talk about this so so let me talk about a few of these with containers it was you know containers are great they're very very fast you can whip up a container you know run the app tear it down really super quickly that works great virtualization is also great the nice thing about virtualization is is they're very very secure and secured by a combination of hardware and software and the reason that that this works this way is because in containers you know you have many many containers running on effectively one kernel so if that kernel is compromised you the entire batch of containers is compromised with virtualization that's not the case so we started a project Intel Clear containers where you no longer have to pick between the security of virtual machines and the and the speed of and the speed of containers where we marry all these two together where we do enable that and this is some of the work that you probably have started seeing in integrated in dockers in Kubernetes and so on and we will continue that work but this is something that I do encourage you to start participating in another great technology that we started and this is this is going back to the car example so the car example that the car is moving is a moving data center so that needs to be orchestrated and you need to have a similar software defined infrastructure but you need to have that software defined infrastructure without a PhD in cloud software and without without a dozen system administrator to run it this needs to be very very fast very secure and and and very low overhead in terms of set up and tear down and so on and and this is what this is what we what we are starting with with Chao which is a very lightweight orchestrator specifically built with with with lightweight usages like a moving data center like edge computing as in mind to to be able to support that and another item that we're barely thinking about and quite frankly we don't even know if it's possible the the idea of if you look at critical systems today whether in in cars or in whether in on aeroplanes or any kind of critical system there is the safety critical or as the functionally safe operating system functionally safe low-level software and and typically this kind of software is built like from the ground up with that specifically specific safety properties in mind so and it's very very small very very small in some cases these operating system microcontrollers is nothing by by a hand inspected do loop so with the way we're proposing to have a full-featured operating system that is able to run on a car or or a hypervisor or a hypervisor that would run underneath a car that is defined the software an entire software defined cockpit these things needs a functionally safe operating system that is full-featured so the idea so with a bunch of our engineers have been working on the idea of can we make Linux a functionally safe operating system we colloquially we call this safety critical Linux we actually think it's possible but it's going to take a lot of work to to create a function safety critical Linux or a functionally safe Linux and a functional safe hypervisor some of it is going to be completely new and some of it is really creating something you know evolving some of the existing operating systems into doing those type of functions time sensitive networking again this is something that we've been working with organization like Avenue and it's really simple in the real-time or asses you know you expect real-time response on the same machine but what happens when you're on the network when you are expecting some kind of a or at least predictable response from an interaction with the cloud so all of these type of projects and there is by the way like 20 other more I can talk for the next hour about all sorts of projects that you should start seeing from us and we want you know everybody start participating in but fundamentally the same idea looking at the end-to-end cloud to edge with all the networking and everything inside and and that is really what we want to focus on and those type of new things that we want to focus on anyway thank you very much for your good listening and we'll we'll see you around the conference thanks