Cluster Resources at a Glance
NYU HPC maintains and manages several compute clusters for use by NYU researchers. For historical usage information, please see the cluster utilization charts.
| Union Square (USQ) | Bowery | Cardiac |
---|---|---|---|
Vendor/Model | Dell Poweredge1950 | Dell | SUN |
Network | IB DDR | IB QDR IB | IB DDR |
Operating System | RHEL 5.1 | RHEL 5.1 | RHEL 5.1 |
# of Login Nodes | 2 | 4 | 1 |
# of Compute Nodes | 73 | 230 | 79 |
# of CPU Cores | 584 | 2,528 | 1,264 |
Total Memory | 1.50 TB | 8.95 TB | 2.47 TB |
Theoretical Peak Performance | 5.44 Teraflops | 28.23 Teraflops | 11.63 Teraflops |
CPU | Intel Xeon | Intel Nehalem | AMD Barcelona |
Preferred use | Serial | Serial/Parallel | Serial/Parallel |
Storage Resources at a Glance
NYU HPC manages a three-tiered storage system for use by the NYU research community. For your disk usage information, please see section Managing Data.
Mountpoint | Size | FS Type | Backup | Availability | Variable | Value |
---|---|---|---|---|---|---|
/home | ~1TB | ext3 | Yes | All nodes (login, compute) through NFS over ethernet. | $HOME | /home/$USER |
/scratch | 301TB | Lustre | No | All nodes (login, compute) through Lustre over Infiniband. | $SCRATCH | /scratch/$USER |
/archive | 102TB | ZFS | Yes | Login nodes only through NFS over ethernet. | $ARCHIVE | /archive/$USER |
PBS Script Generator
Front-Line HPC Consulting