Use Cases

Examples By Category

Distributed Python and interactive Jupyter

Easy-to-use and HPC-optimized data exchange between applications

Observability for applications and workflows

AI/HPC workflows across systems and sites

Data loading and resilient training and inference

Process a large dataset in parallel

Orchestrate executing a serial executable with many different arguments

Orchestrate a parameter sweep for an MPI application

Train with PyTorch across many GPUs

Track application state to enable fault tolerance and recovery

Run on a collection of servers or a supercomputer

Use a Jupyter notebook with Dragon

Running functions and processes that use specific GPUs

Visualize and analyze system and custom metrics

Develop a workflow that puts it all together