You are here

Using the CRAY XT-6m

The CRAY supercomputer at UDE has been in service since the year 2010 and is suitable of massively parallel computing. Most modern scientific codes are intended to run on more than one core. If your code can be executed on a large number of cores (>= 48) at the same time, the CRAY might be of service.

Working code:

  • VASP (5.3.3)
  • LAMMPS (09/2013)

The binaries and job files can be found in the folder /imports/fs/opt/cray/

If you want to try a different code or/and have problems compiling it on the supercomputer, just come to my office (S05 V06 E86).

Queues:

The maximum number of jobs is limited to 15 per user and to 8 per queue. For testing purpose there is a special queue "testq". This queue is limited to 96 cores and the maximum walltime is one hour.  The other queues are c528t96h, c1kt48h and c2kt24h. The first number gives the maximum number of cores and the second number the maximum walltime in hours.

WARNING:

  • If your code is only executable on a single core or does not parallelize well, do not use the CRAY.
  • Running jobs with less than 12 cores per job is a waste of resources.
  • Most codes do not work properly if they are not compiled at the supercomuter.

 

Theme by Danetsoft and Danang Probo Sayekti inspired by Maksimer with minor local changes