Multi-Instance GPU (MIG) is a new feature of NVIDIA’s latest generation of GPUs, such as A100, which enables (multiple) users to maximize the utilization of a single GPU, by running parallel compute workloads as if there were multiple smaller GPUs. In this training, we will share some best practices about how to run multiple deep learning applications in parallel on a single A100 using MIG, etc. After this training, you will be able to use MIG to run multiple deep learning workloads in parallel or allow multiple users to share an A100 GPU, in order to improve the utilization of a single A100 or extend this idea to multiple GPUs on a DGX A100