 Actually, you should be my colleague who will do this presentation, but it's absent today, so I will do the presentation for you. We will talk about Lightweight Age Computing Eluit. Why do we have Lightweight Age? Now it's all digitalized, light digitalized industries. You will create a huge flow of data. If you upload all this data to cloud, then you need to have very high bandwidth. For connected vehicles, our medical services require very low latency, but if you upload all this to the cloud, then it will create higher latency. If you can do this on the edge and it will create lower latency, for the cloud, it will provide storage and large scale computing, but for age, it will focus on real-time low latency. And then we have this Lightweight version. Here, a research report shows that age computing has entered industry high-speed developing period. And this very famous investment research institute says that it will develop following a spiral route. Previously, you will have one big machine in your lab to process some tasks, and then in the 1980s, you have all the species. And now, you have the mobile internet and then by 2020, it will actuate in the new era of intelligence. And by 2020, 50% of the data will be transferred from the data center to the age end. Now, it's experiencing a very important construction period. It's developing very rapidly. Now, we have more and more age devices like artificial intelligence, microservices, standardized protocol, etc. All of this will increase the complexity of computing. You have different kind of hardware devices. By open source, we can reduce the complexity of management and integration. In this open source community, all the companies, vendors, you can work together to work on the code. You will do some testing in the community, and this can better manage the complexity. ICT infrastructure now is open to many industries and that will promote the digitalization transformation of many industries. And finally, you have the results as the right graph shows here. It will accelerate the business transformation. And in this case, the industries can invest more on ICT. Before this, you might have, for example, a patch, open stake, foundry, etc. And a kernel and age stack, they are more like end-to-end integrated community. And FN is like an M-brother to cover all this and to solve this fragmented issue. It will cover the high-value scenarios like enterprise, industry, end-to-end and end-to-end testing, validation, etc. And also age computing OS. On the right, it's the structure of LFH. On the top, you have the board. Below that, you have the technical adversary. Age and home age. All those are under the LFM banner. LFH now has 63 members including Huawei. Under the LFH, there are five communities. I only list three here. And there will be more companies that will join this umbrella. So this is the first release of a kernel incubating project. And it has moved calculating computing capability from the center to the edge. And here is a major scenario, such as the Huawei edge lights. Here is the structure of the alien. Here is the major architecture. So on the edge side, there will be a local data center. So on the center, there will be a role to manage the alias nodes and also the crucial data management. These are edge sites. For example, the gateways with the sensors and devices. So this is a simple VRAR monitoring device. So it's an end-to-end design. So it supports the structures you can see here. And there's a lightweight virtual structure using a large scale operating system. Because on the edge, there's only an LT gateway if you use a large operator. And also there's a lightweight pass. It complements the lightweight. And moving to the upper level, there's tailoring on the lightweight. And also here is a scenario application. So the first phase, the phase one alias has been released. Supporting the edge cost of management and monitoring. Also support from this cluster. This is the structure image. The edge pass. Well, Coopee Edge is also a lightweight edge. This is an Elliott lightweight video monitoring plan. Here we have some problems on the edge side. The resources are limited. But we still need to process the data. Not all the video traffic are set to the cloud. So our solution is to trim the edge knots container and make it lightweight. So if there are other applications and demands, there will be other customized trimming. This is the Elliott video monitoring topology. So the manager is on the cloud or on the data center including three components. First, second, and third. And on the edge knot, there is a nice part. Such as the face recognition function and also the TV device. First, we need to integrate these knots and make them correspond with the lower level video streaming. Then we can do the connection here. So after the face recognition, the data will be set. These knots will be refreshed. And the matches meet the standard. And the data will be sent to the central cloud and stored. For example, if there's on a roll, there are 7 cameras. So after the data is processed on the edge, these data will be released and sent to the Elliott manager into database. If someone has one spot, monitored by one camera on a road, then at the time he was spotted at another by another camera, then the matching can be realized through the data storage. And then we can track the path that this person walks on the road. And also, using Prometheus, we're looking to resource monitoring on ALS bus. So here is your UI. We can monitor the resource usage condition. So at last, I want to say these join Elliott's and now contribute to this new platform. So the key takeaways of Elliott's includes we focus on the lightweight edge platform and also IOT gateway. And also, we have very good commercialization pathways. So welcome for your joining. And so we're going to do commercialization together. So since here I represent the speaker today to do the speech. So if you have any questions, please send the questions to this email address. Thank you.