What you need to know
Once you’ve been awarded your allocation, you will need to interact with the ALCF’s PBS Professional job scheduler and launch your code on your allocated system(s).
At a high level, getting computational tasks run on an HPC system is a two-step process:
- You request and get allocated on ALCF resources at the node level on one or more of the systems. We use PBS Professional as our job scheduler / workload manager.
- You execute your tasks on those ALCF resources. This is accomplished in your job script by interacting with various system services (MPI, OpenMP, the HPE PALStask launch system, etc.)
Running Jobs on ALCF Systems
You can learn more about running jobs on each of our specific systems in our documentation.