The NVIDIA vGPU Device Manager manages vGPU devices on a GPU node in a Kubernetes cluster. It defines a schema for declaratively specifying the list of vGPU types one would like to create on the node. The vGPU Device Manager parses this schema and applies the desired configuration by creating vGPU devices following steps outlined in the NVIDIA vGPU User Guide. Currently, this image is deployed as an operand of the NVIDIA GPU Operator and is only to be used in the context of Kubernetes.

Features

• Declarative Configuration: Defines a schema for specifying vGPU types.
• Schema Parsing: Parses the schema to apply the desired vGPU configuration.
• vGPU Creation: Creates vGPU devices following the NVIDIA vGPU User Guide.
• Kubernetes Integration: Deployed as an operand of the NVIDIA GPU Operator for Kubernetes clusters.

Benefits

• Simplified Management: Simplifies the management of vGPU devices in Kubernetes.
• Flexible Configuration: Allows flexible and declarative specification of vGPU types.
• Efficient Resource Utilization: Ensures efficient utilization of GPU resources.
• Seamless Integration: Integrates seamlessly with Kubernetes and the NVIDIA GPU Operator.