HPC
HPC copied to clipboard
Update Vermilion Documentation to include questions that came up in Vermilion User Meetings
Questions/things to clarify include:
- adding documentation for building GPU apps on Vermilion
- Some questions on terminology, like "virtualization". Maybe we can clarify some terms
- General advice about which jobs should be run on Vermilion and which ones would be better off on Eagle/Swift/Kestrel (how to split up jobs between systems for projects with allocations on multiple systems)
- Two people reported not being able to log in to Vermilion using ssh vermilion.hpc.nrel.gov, but were able to log in with ssh [username]@vermilion.hpc.nrel.gov. Might be worth making a note of this in the documentation or looking into why this is the case
- add information about allocation charges to the partitions table
- add info about filesystem memory limits to documentation and guidance about where to run jobs
Allocation charge factors addressed in PR #481