You can now seamlessly run machine learning and high performance computing applications by scheduling Amazon Elastic Container Service (ECS) tasks with Amazon EC2 GPU instances in the same manner that you do CPU and Memory. You can also pin these GPUs instances to containers for workload isolation.