Removal of broadwell nodes from Hydra

The broadwell partition will be removed from Hydra on Sunday 16 of June at midnight. Only jobs that can finish before that date will be able to run on the broadwell partition.

The removal of the 27 nodes with Intel Broadwell CPUs is justified by their age, as they are more than 6 years old and have low energy efficiency by current standards. Moreover, as shown in our recent usage report “Hydra Overview 2023”, the broadwell partition has been in low demand for the past year and the capacity provided by it is not needed to handle the job queue in Hydra.

Please review your job scripts and replace any existing mention to the broadwell partition in them, specifically the --partition setting for Slurm. In general, jobs should continue to work without any specific partition defined. If you need to target one partition, the best alternatives are the recently added zen4 nodes using AMD Genoa-X CPUs or, if you want an Intel based solution, the skylake partition using Intel Skylake CPUs.

This change also affects the notebook platform, which will no longer offer the option to launch notebooks on the broadwells.

Note

The removal of the broadwell nodes does not affect the broadwell_himem partition in Hydra. This special system with 1.5 TB of RAM memory will continue to be available.

You can contact us at VUB-HPC Support with any comments or questions about this change.