GPU resource using method and device and storage medium

Abstract

The invention relates to the technical field of high-performance computing, and provides a GPU resource using method and device and a storage medium. The GPU resource using method comprises the stepsthat a container management system creates a scheduling task according to a calculation request submitted by a client; the container management system judges whether the idle GPU resources in the cluster meet the resource requirements specified in the instruction for creating the GPU container or not; if the requirement is not met, the container management system suspends the scheduling task firstand then executes the scheduling task until the requirement is met; and when the scheduling task is suspended, the container management system does not create the GPU container. In this method, the container management system queues the scheduling tasks according to the use conditions of the GPU resources; therefore, for each computing task, the GPU resources can be used exclusively, the GPU resources used by the computing task are not shared with other computing tasks, and the execution progress and the computing result of the computing task can be carried out according to a plan and cannotbe affected by other computing tasks.

Description

technical field [0001] The present invention relates to the technical field of graphics processor (Graphics Processing Unit, GPU for short), and in particular, to a GPU resource usage method, device and storage medium. Background technique [0002] Container technology (for example, Docker) is a new type of virtualization technology. Using container technology, developers can easily deploy and run applications on containers, and easily realize automatic installation, deployment and upgrade of applications through configuration files. It is convenient to separate the production environment and the development environment without affecting each other. [0003] A GPU container (eg, Nvidia-Docker) is a container that can use a GPU device. At present, in the field of scientific research, a large number of computing tasks (for example, deep learning tasks in the field of artificial intelligence) use GPU containers as the operating environment in order to use high-performance GPU ...

AI Technical Summary

Problems solved by technology

However, for some specific computing tasks, being interrupted by other tasks during the running process (due to switching to the time slice of other computing tasks) may delay its execution progress and / or cause the calculation results to be unavailable, so it is not suitable to use virtual Allocate GPU resources in the cluster in a standardized way

Images