Node Types and Subclusters
SciClone and Chesapeake are clusters of many individual computers ("nodes"), so before you can run a job on SciClone or Chesapeake, you must decide on which particular member computers your job will run (often one or more groups of similar computers, called subclusters). The table below is intended to assist you with this decision, after which you can learn how to submit jobs using the Torque or Slurm batch system.
Please contact the HPC group for advice if you are unsure of the most appropriate type of node for your particular application.
Main-campus/Sciclone cluster
- kuro - MPI parallel jobs requiring at least 64 cores (Slurm)
- astral - shared memory GPU cluster w/ 8x Nvidia A30 GPUs, 24GB (Slurm)
- gulf - 2 High memory Nodes & 4 GPU nodes cluster w/ 2x Nvidia A40 GPUs, 48GB (Slurm)
- bora - Main MPI/parallel cluster (Torque)
- hima - Main shared memory cluster with some GPUs (Torque)
- gust - Main cluster for large memory / shared memory jobs (Slurm)
- femto - Currently reserved for Physics and VIMS use (Slurm)
- vortex-alpha - Currently exclusive to AidData (Torque)
VIMS campus/Chesapeake cluster
- James - MPI/parallel cluster (Slurm)
- Potomac - Serial /shared memory / small parallel jobs (Slurm)
- Pamunkey - Shared memory only / exclusive to bio/bioinformatics calcs (Slurm)
Main-Campus ("SciClone") Cluster |
||||||
|---|---|---|---|---|---|---|
| Name |
Cores per node / Threads per node |
Memory (GB) |
/local/scr | Network | OS | Deployed |
|
SciClone Kuro cluster (3008 AMD Zen 4 cores) Batch System: Slurm |
||||||
|
kuro.sciclone.wm.edu |
32 | 384 | 1.8 TB |
10 GbE HDR IB |
Rocky 9.2 | 2024 |
|
ku01-ku47 |
64 | 384 | 980 GB |
1 GbE HDR IB |
Rocky 9.2 | 2024 |
|
SciClone Gulf Cluster (16 CPU cores w/ 512G Mem + 32 CPU cores w/ 2 Nvidia A40 GPUs) Batch system: Slurm |
||||||
|
gu01-gu02 |
16 | 512 | 1.8 TB |
10 GbE HDR IB |
Rocky 9.3 | 2024 |
|
gu03 - gu06 (each w/ 2x Nvidia A40 GPU (48GB) |
32 | 128 | 1.8 TB |
10 GbE HDR IB |
Rocky 9.3 | 2024 |
|
SciClone Astral Cluster (32 CPU cores + 64 CPU cores/8 Nvidia A30 GPUs Batch system: Slurm |
||||||
|
astral.sciclone.wm.edu |
32 | 256 | 63 GB |
10 GbE HDR IB |
Rocky 9.2 | 2024 |
|
as01 (has 8x Nvidia A30 GPU (24GB) |
64 | 512 | 14 TB |
10 GbE HDR IB |
Rocky 9.2 | 2024 |
|
SciClone, Bora and Hima subclusters (1324 Xeon "Broadwell" cores / 2648 threads) Batch system: Torque |
||||||
|
bora.sciclone.wm.edu (front-end/login) |
20 |
64 | 10 GB |
10 GbE, FDR IB |
CentOS 7.3 | 2017 |
| bo01-bo55 | 128 | 524 GB |
1 GbE, FDR IB |
|||
| hi01-hi07 |
32 |
256 | 3.7 TB [2] |
1 GbE, QDR IB |
||
|
SciClone Gust subcluster (256 EPYC "Zen 2" cores) Batch system: Slurm |
||||||
|
gust.sciclone.wm.edu (front-end/login) |
32 | 32 | 64 GB |
1 GbE, EDR IB |
Rocky 9.3 | 2020 |
| gt01-gt02 | 128 | 512 | 670 GB |
10 GbE, EDR IB |
||
|
SciClone, Femto subcluster (960 Xeon "Skylake" cores) Batch system: Slurm |
||||||
|
femto.sciclone.wm.edu (front-end/login) |
32 | 96 | 10 GB |
1 GbE, EDR IB |
CentOS 7.6 | 2019 |
|
fm01-fm30 |
32 | 96 | 2 TB [3] |
1 GbE, EDR IB |
CentOS 7.6 | 2019 |
| VIMS CLUSTER | ||||||
|
Chesapeake: Potomac subcluster (360 "Seoul" and compute cores) Batch system: Slurm |
||||||
|
chesapeake.hpc.vims.edu (front-end/login) |
12 | 32 | 242 GB |
10 GbE, QDR IB |
CentOS 7.9
|
2014 |
| pt01-pt30 | 12 | 32 | 242 GB |
1 GbE, QDR IB |
||
|
Chesapeake: James and Pamunkey subclusters (540 "Skylake" Xeon 128 "Abu Dhabi" Opteron compute cores) Batch system: Slurm |
||||||
|
james.hpc.vims.edu (front-end/login) |
8c/ 16t |
32 | 10 GB |
10 GbE, EDR IB |
CentOS 7.5
|
2018
|
| jm01-jm27 |
20 |
64 | 1.8 TB |
1 GbE, EDR IB |
||
|
pm01-pm02 |
64 |
256 | 1.3 TB [2] |
10 GbE |
CentOS 7.3 |
2016 |
|
choptank (/ches/data10) |
12c/ 24t |
64 | 800 GB |
1 GbE, QDR IB |
RHEL 7 |
2016 |
|
rappahannock (/ches/scr10) |
32 | 176 GB |
10 GbE, QDR IB |
2014 |
||
-
va01-va10are on a separate InfiniBand switch and have full bandwidth with each other, but are 5:1 oversubscribed to the main FDR switch. - Usually a node's local scratch filesystem is a partition on a single disk, but the local scratch filesystems on Pamunkey and Hima nodes are faster (~300 and ~400 MB/s) six- and eight-drive arrays, respectively.
- The femto nodes are equipped with a 2TB SS Disk for fast reads and writes.
