The Tango cluster has 7 big memory nodes that can be utilised for jobs requiring up to 1TB RAM.

Using the SLURM Workload Manager, jobs specifying more than 256GB will run on these nodes .This eliminates the need for a separate queue for big memory jobs which in turn should reduce queue times.

Note: Please see Scratch on Tango for details on how to use the scratch directory on Tango.



Did you find this information useful? Share your feedback here.