G100: maintenance completed

  1. /
  2. HPC Center news
  3. /
  4. G100: maintenance completed

Dear User, 

this is to inform you that the maintenance of Galileo100 has been completed and the cluster is now back in production.
We would like also to inform you that a new partition has been added to Galileo100: g100_usr_pmem.
This partition is part of the already existing g100_usr_prod and is characterised by a maximum memory per node of  375300MB, with additional features that will be better described in following updates. Instructions and best practices for how to use efficiently these areas will arrive soon. In order to request this partition, you need to specify the Slurm directive:
#SBATCH -p g100_usr_pmem
in your batch script.
As for the g100_usr_bmem partition with nodes with a maximum memory of up to 3TB, they have been reduced in number and separated from the standard g100_usr_prod partition, so they now have to be requested explicitly:
#SBATCH -p g100_usr_bmem.
Best regards,
HPC User Support @ Cineca