Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

Table of Contents
maxLevel2

(updated: May 2022 )

...

hostname:                  login.dgx.cineca.it

...

Since DGX filesystems are based on GlusterFS the usual unix command "quota" is not working. Differently, for other CINECA clusters, cindata command is not yet available. Use the local command cindata to query for disk usage and quota ("cindata -h" for help):

> cindata

For information about data transfer from other computers, please follow the instructions and caveats on the dedicated section Data storage or the document  Data Management.

...

SLURM

partition

QOS

#cores/#GPUs per job

max walltime

max running jobs per user/

max n. of cores/GPUs/nodes per user

Priority

notes

dgx_usr_proddgx_qos_sprod

max = 32 cores (64 cpus) / 2 GPUs

max mem=245000MB

48 h

1 job per user

32 cores (64 cpus) / 2 GPUs

30


normal (noQOS)

max = 128 cores (256 cpus) / 8 GPUs

max mem=980000MB

4 h

1 job per user

8 GPUs / 1 node per user

40


dgx_usr_preemptdgx_qos_sprod

max = 32 cores (64 cpus) / 2 GPUs

max mem=245000MB

48 h(no limit)1free of charge / your jobs may be killed in any moment if a high priority job requests for resources in dgx_usr_prod partition

normal (noQOS)

max = 128 cores (256 cpus) / 8 GPUs

max mem=980000MB

4 24 h(no limit)1free of charge / your jobs may be killed in any moments if a high priority job requests for resources in dgx_usr_prod partition

...

On each node of DGX cluster singularity is installed in the default path. You don't need to load singularity module in order to use it, but we have created it to provide some examples to users via the command "module help tools/singularity". If you need more info you can type singularity --help or visit Singularity documentation web site.

...

Here you can find more examples about pytorch.

Build your own container

Please refer to the relative section of this page in our User Guide.