Image: /cuda-vector-add:v0.1ĭeploy the container group with the az container create command, specifying the YAML file name for the -file parameter. Learn more about working with public images. To improve reliability, import and manage the image in a private Azure container registry, and update your YAML to use your privately managed base image. The following example uses a public container image. For example, you can use the following base images for your Dockerfile: We support up through CUDA 11 at this stage. It does not include the time to deploy the container group.ĬUDA drivers - Container instances with GPU resources are pre-provisioned with NVIDIA CUDA drivers and container runtimes, so you can use container images developed for CUDA workloads. The duration is calculated from the time to pull your first container's image until the container group terminates. Pricing - Similar to container groups without GPU resources, Azure bills for resources consumed over the duration of a container group with GPU resources. This is due to the additional time to provision and configure a GPU VM in Azure. Things to knowĭeployment time - Creation of a container group containing GPU resources takes up to 8-10 minutes. To request an increase in an available region, please submit an Azure support request. The default CPU limits for the P100 and V100 SKUs are initially set to 0. These values are currently larger than the CPU and memory resources available in container groups without GPU resources.ĭefault subscription limits (quotas) for GPU resources differ by SKU. When deploying GPU resources, set CPU and memory resources appropriate for the workload, up to the maximum values shown in the preceding table. Each SKU maps to the NVIDIA Tesla GPU in one the following Azure GPU-enabled VM families: SKU

To use GPUs in a container instance, specify a GPU resource with the following information: Support will be added for additional regions over time.Īdditional limitations: GPU resources can't be used when deploying a container group into a virtual network. In preview, the following limitations apply when using GPU resources in container groups.

If you would like to use this sku for your production container deployments, create an Azure Support request to increase the limit.

Due to some current limitations, not all limit increase requests are guaranteed to be approved.

Cuda driver