Office of Research Computing

We had minor storage issues earlier that caused one storage node to take over for another. This is usually a successful operation that you wouldn't notice, but this time some client compute nodes did not reconnect to the storage for some reason. Some jobs may have died and we are working to figure out why that happened. Last Updated Thursday, Feb 12 04:17 pm 2026

Job Script Generator

Video Tutorials

Parameters (video tutorial)
Limit this job to one node:
Number of processor cores across all nodes: #nodes * #cores
Number of GPUs: Very limited number of GPUs available.	Only use this if your code actually utilizes GPUs.
Memory per processor:
Due to resource requirements and scheduler software limitationsyou *must check the bigmem feature* in the list of features for your job to run on a big memory node.Required for all jobs with ppn>16 or memory per node > 48 GB
Walltime:	hours mins secs
Job is a test job:
Job is preemptable:
Use Inifiband:
I am in a file sharing group and my group members need to read/modify my output files:
Group name (case sensitive):
Need licenses?
License names and counts:	Name Count Name Count Name Count
Job name:
Receive email for job events:	begin end abort
Email address:

Job Script Generator

Video Tutorials

Features

Job Script