Computility Grid

With the rapid development of AI big models, the demand for computing power is increasing day by day. As an important infrastructure supporting these computing tasks, computing power networks contain various heterogeneous resources, such as computing nodes with different performance, storage devices, and network bandwidth. The effective scheduling of these heterogeneous resources is directly related to the efficiency and quality of computing task execution. By reasonable scheduling of heterogeneous resources, the advantages of various resources can be fully utilized to achieve fast and efficient execution of computing tasks. For example, for tasks that require high computing power, they can be scheduled to computing nodes with stronger performance; For tasks that require a large amount of data storage, the space of storage devices can be fully utilized; At the same time, it is also necessary to consider the allocation of network bandwidth to ensure fast data transmission.

This project will study efficient scheduling algorithms for heterogeneous computing power (such as CPU, GPU, FPGA); Unified scheduling architecture for supercomputing and intelligent computing; Task orchestration and scheduling mechanisms for cross domain and cloud edge collaboration.

Wei Dong
Wei Dong
Professor

Wei Dong is currently a full Professor at the College of Computer Science, Zhejiang University.

Related