GPU cloud server implementation method based on OpenStack architecture

The invention provides a GPU cloud server implementation method based on an OpenStack architecture, and belongs to the technical field of cloud computing. The method comprises firstly, selecting a GPUphysical machine with specific resource configuration; erecting the GPU physical machine to a computing node of an OpenStack; configuring host aggregation and cloud host specification Flavor corresponding to the GPU physical machine according to the GPU card; configuring a scheduling algorithm, a control node and a computing node of the OpenStack to meet the passthrough of the GPU to a virtual machine, namely a GPU cloud server; and enabling the GPU cloud servers to realize mounting and unloading of GPU cards between the cloud servers on the same GPU physical machine through a PCI hot plug technology. 本发明提供一种基于OpenStack架构的GPU云服务器实现方法，属于云计算技术领域，首先选择资源特定配置的GPU物理机，把GPU物理机上架到OpenStack的计算节点，并按照GPU卡配置GPU物理机对应的主机聚合和云主机规格Flavor，然后配置OpenStack的调度算法、控制节点和计算节点来满足GPU卡透传(passthrough)给虚拟机即GPU云服务器，GPU云服务器通过PCI的热插拔技术实现GPU卡在同一GPU物