So let's take a few moments and talk about the visual cloud providers. In this area we typically break it up into two basic industries, communications service providers and their use cases first. And here we find that this is actually the ingress point into the deeper part of the network. So functions like broadcast quality and video on demand delivery are some of the key areas there, probably primary use cases. The through the network or the over the top video delivery would see demand placed on that comm service provider for their content delivery so that they can offload congestion in the network by proactively moving some of that media on demand. And potentially then, because of that, accessing analytics so that we can do things that we spoke of before like add insertion as well as the immersive media, where the latency of the round term delay is going to be very critical. We may not be able to pay the latency costs of moving it deep into the network and then into the cloud service provider at that point. So these are some of the critical elements that we would find some of use cases inside the communications service providers. And there are a few more listed here that we didn't get a chance to talk to. You can take a look at those on your own time. And the second, obviously then, is the cloud service provider itself. So here we're saying that the cloud service provider is probably a very large type data center operator, who you can think of, who has resources at the far edge of the network deep into it behind a communication service provider who is providing that. And obviously, you know the through the network itself, the source of those videos is oftentimes in that cloud service provider area. So we see that as a potential for optimization of some of the functionality, particularly when we talk about the transcoding is going to deliver that content into the network and maybe we want to transcode it at that source before we deliver it deep into the network so that we have efficiencies then that we drive into it. Online gaming is a great area that we see as a use case in here. And also, they can certainly take advantage of the immersive media and the media analytics, given their ability to identify the user and associate that with maybe some of their past visibility or past viewing content so that they can offer them things, like well, if you're interested in A you'd also be interested in a derivative of A. And here's this list of the derivative of A type of thing to drive more traffic into their area. So these are all some of the great use case examples where we see the visual cloud and then the visual computing underneath it as possible use cases in that cloud environment. So what's that flow look like then? And I sort of have waved my hands a little bit at the time when I talk about information coming in from that source, from that edge, where the content is created or captured. So you know, we think about that, capturing. Cameras doing that work, or creative. You're doing 3D rendering, or you're creating some type of image that's going to be used as an AR or VR type of image. And then that information needs to flow through from an edge point at capacity and at speed into that network. So we need a low a low latency, high ping with network in order to deliver that. Again, it's not a single point, but we need to think about this from a global standpoint. Typically at a very high quality. We don't lose any information. We've mention the lossy and lossless types of compression, but we want that to be as lossless as possible for very high quality, very high capacity, very high density. And for that one we look at Intel technology. These are great applications both for the Intel Xeon SP line of processors and FPGA for acceleration types of workload. Moving that data deeper into the network, then we've got production processing, where we can bring compute resources, production manipulation, video data analytics onto that content. And then finally distribution, and processing, and storage. And as we work further to the right, that content needs to flow back out in order for it to be to be consumed, to be useful. And again, here we encounter the needs of the network to provide that low latency and high bandwidth and finally that edge deliver processing where we may be additional buffering for that CDN type of operation discussion that we had where we've got pre-created content and we can rely on that. And certainly, it doesn't work so well when you think about the live streaming. But at that point, we also may have that transcoding, that encoding that goes on as you can have different consumers, consuming that information on different devices, whether it's for 4K, or HD, or even a low density as are consuming on their device. And as you get closer to that side of the edge, graphics capabilities, as you might find, for example, in the Intel Xeon E3 or the QuickAssist video, or even Intel [? prographics ?] consumption are good candidates as you start looking at delivering that as it gets multiplexed out for a consumption standpoint. So that's the sort of the lifecycle, as you will, then from creation of content into consumption in that visual cloud environment. [INTEL THEME]
