Maintenance today has completed. Brief description of the updates made:
- BeeGFS updates for stability and future OS related updates
- CUDA driver updated to 10.1 and the cuda/10.1 module is now exposed. With the update to 10.1, the newest Tensorflow and PyTorch can be installed directly into a virtual environment. See https://crc.pitt.edu/python for details.
- 56 additional nodes were added to smp/smp (i.e. cluster/partition)
- 2 socket 16 core 3.0 GHz AMD Rome CPU
- 251 GB RAM usable
- If you wish to request these nodes specifically, add #SBATCH --constraint=amd to your job submission script
- An issue with our current Gaussian installation was discovered during testing of these AMD nodes, please update to module load gaussian/16-C.01
- 1 high memory node was added to smp/high-mem
- 2 socket 16 core 2.4 GHz AMD Naples CPU
- 996 GB RAM usable
- If you wish to request this node specifically, add #SBATCH --constraint=amd to your job submission script
If you face any issues with these updates please don’t hesitate to submit a ticket at https://crc.pitt.edu/tickets.
Posted: February 4, 2020
Cluster is down for maintenance.
Posted: January 29, 2020
This is a reminder that we will have a full cluster outage this coming Tues Feb 4 from 8AM to 6PM. Jobs that overlap with this scheduled period will remain queued until completion of maintenance. Please plan you compute jobs accordingly.
Posted: January 7, 2020
Our next quarterly cluster maintenance will be Tues Feb 4 from 8AM to 6PM. This cluster-wide downtime will allow us opportunity to make network updates, upgrade our Application VM infrastructure, and to perform general housekeeping to sustain smooth operations at the Center.
Websites that are hosted on our Application VM server will be inaccessible during this maintenance period. Please plan accordingly.