Mix scheduling method facing central processing unit (CPU) and graphic processing unit (GPU)

Abstract

The invention provides a mix scheduling method facing a central processing unit (CPU) and a graphic processing unit (GPU), which comprises the steps of S1 reading scheduling strategy configuration and priority level configuration parameters when a scheduler is initialized; S2 reading information in a job scheduling system in each scheduling circulation; S3 conducting priority level calculation by the scheduler according to operation resource request information and scheduling strategies, and determining the priority level of each operation and ranking in a descending order; S4 conducting operation scheduling according to the scheduling strategies and obtaining ranking sequences obtained in the S3; S5 sending operation start requests to a scheduling system according to scheduling results of the S4; and S6 keeping dormant for a period time, and then returning S2 to conduct a next circulation. The mix scheduling method sets relative relation between the CPU and a non-GPU through customized weight, CPU resource situation is judged during scheduling, and accurate scheduling of GPU operation is achieved.

Description

technical field [0001] The invention relates to job scheduling in high-performance computing, in particular to a hybrid scheduling method oriented to CPU and GPU. Background technique [0002] In recent years, with the popularity of GPU devices and the migration of applications to GPUs, more and more high-performance computing clusters have begun to use GPU devices to support large GPU applications. This puts forward new requirements for the traditional job scheduling system. The resources managed by the traditional job scheduling system are mostly operating system resources (such as nodes, memory, CPU, etc.), and the GPU is not described as an available resource in the scheduling system and participates in scheduling decisions, so there are often two aspects. Questions: One is how to describe GPU resources and GPU resource requests; the other is how GPU applications coexist with traditional parallel applications (MPI, OpenMP, PThread) to ensure the rational use of system r...

AI Technical Summary

Problems solved by technology

The resources managed by the traditional job scheduling system are mostly operating system resources (such as nodes, memory, CPU, etc.), and the GPU is not described as an available resource in the scheduling system and participates in scheduling decisions, so there are often two aspects. Questions: First, how to describe GPU resources and GPU resource requests; second, how to coexist GPU applications with traditional parallel applications (MPI, OpenMP, PThread) to ensure the rational use of system resources