Running jobs
Getting work onto the compute nodes efficiently.
| Title | Description |
|---|---|
| Workflow managers | Workflow managers are useful for reproducible workflow, especially when you have complicated pipelines with complicated data dependencies. |
| Embarassingly parallel jobs | There are many ways to launch a bunch of embarassingly-parallel (TODO: def) tasks on an HPC. |
| Oversubscription |
On an HPC, it is quite often you’d run into a situation that you need to think about restricting the number of threads in multithreading scenario where if treated not… |
No matching items