Head of Live Video Technology
Track: High Performance & High Concurrency Server Side
2019/12/13 14:00 - 17:55
From 4K multiple real-time transcoding, massive short video real-time coding, How to design the architecture on the server for all kinds of application scenarios? How to maximize the value of the existing software and hardware? How to achieve high cost-effective by heterogeneous computing?
Topics in the track
HUYA has been trying to break through the limit of the live broadcasting industry, such as the first 10M Blu-ray of the live broadcasting industry, large-scale CDN P2P, and other technologies. With the development of 5G technology, a large number of different applications will be created in the upcoming future as new applications and business models will be introduced, which require better and higher network performance. Edge computing technology is one of the core technologies to solve the diversified network requirements brought by different applications. It increases the computing ability in the server room which close to the Client, backward the computing power, achieves a qualitative leap regards to user experience. HUYA uses the self-built CDN converged network, and the large-scale, highly flexible edge computing resources, by opening to the streamer end and the audience side, achieves complex computing logic, provides higher HD videos and satisfies diversified product requirements. Edge computing application system: UDP video uplink, HUYA self-built p2p, 3D virtual live broadcast, multi-person connecting broadcasting, and outdoor HD live broadcast
As video gradually becomes the mainstream expression of content for users, the video processing capability of the server also faces greater technical challenges as an important link to protect the user's production and consumption experience and reduce the operating cost of the platform. This presentation will bring you the best practices of large-scale video cloud processing systems in the process of high-availability guarantee, performance optimization and cost optimization in the process of ensuring rapid iteration of services by Meitu's large-scale video cloud processing system.