About Multi-Instance GPU (MIG)
The partitioning option for NVIDIA Multi-Instance GPUs is now supported. For better utilization, you can partition your GPUs into smaller slices.
By default, a workload would use the entire GPU, but if the GPU is partitioned (or sliced) to
smaller pieces, then the workload would just use a smaller portion of the GPU and multiple
workloads could be running on the same GPU.
