Hands on One of the biggest headaches associated with AI workloads is wrangling all the drivers, runtimes, libraries, and other dependencies they need to run.
Deploy an application to Amazon ECS with EC2 | Dock workers | ECR | Fargate | Load Balancer | AWS project
This is especially true for hardware accelerated tasks where if you have the wrong version of CUDA, ROCm or PyTorch there's a good chance you'll be scratching your head staring at an error.
If that wasn't bad enough, some AI projects and apps may have conflicting dependencies, while different operating systems may not support the packages you need. But by containerizing these environments, we can avoid much of this mess by building images that are configured specifically for a task and – perhaps more importantly – deployed in a consistent and repeatable way every time.
And because the containers are largely isolated from each other, you can usually have apps running with conflicting software stacks. For example, you can have two containers, one with CUDA 11 and the other with 12, running at the same time.
