<p>27/04/2021<br />
Dear users,</p>
<p>the cluster Marconi100 has completed the maintenance operation and is now in<br />
production. Please take in consideration the following updates:</p>
<p> </p>
<p>- the version of "module" has been updated to 4.7.0 . Please take a look at<br />
the release notes<br /><a href="https://modules.readthedocs.io/en/latest/NEWS.html#modules-4-7-0-2021-02-19;">https://modules.readthedocs.io/en/latest/NEWS.html#modules-4-7-0-2021-02...</a></p>
<p>- the module of hpc-sdk/2021 now hosts the updated version 21.3, that is<br />
recommended since it provides a better support for GPU offloading. Please<br />
take a look at the release notes<br /><a href="https://docs.nvidia.com/hpc-sdk/hpc-sdk-release-notes/index.html;">https://docs.nvidia.com/hpc-sdk/hpc-sdk-release-notes/index.html;</a></p>
<p>- a new partition "m100_usr_preempt" is in production. The partition allows<br />
for jobs to access nodes usually not available for users, but they can be<br />
killed anytime by jobs of higher priority. For further details please check<br />
our updated Userguide<br /><a href="https://wiki.u-gov.it/confluence/display/SCAIUS/UG3.2%3A+MARCONI100+UserGuide#UG3.2:MARCONI100UserGuide-Submittingbatchjobsforproduction">https://wiki.u-gov.it/confluence/display/SCAIUS/UG3.2%3A+MARCONI100+User...</a>.</p>
<p> </p>
<p>Regards,<br />
HPC User Support - CINECA</p>