Compute Nodes
Please note: depending on server configuration, 2 to 4 CPU cores can be reserved for a nodes' operating system and to ensure SLURM can run without issue.
General use nodes
Partition |
Nodes |
CPU cores |
GPUs |
Group Restriction |
gpu |
19 |
856 |
52 |
none |
long_gpu |
2 |
136 |
6 |
as needs basis, see policy |
cpu |
33 |
4224 |
- |
none |
long_cpu |
5 |
640 |
- |
as needs basis, see policy |
interruptible_cpu |
73 |
4672 |
- |
none, see policy |
Departmental/Faculty nodes
Partition |
Nodes |
CPU cores |
GPUs |
Group Restriction |
nmes_gpu |
8 |
576 |
32 |
er_dpt_nmes |
mathematics_cpu |
2 |
256 |
- |
er_dpt_mathematics |
biomed_a30_gpu |
2 |
96 |
16 |
er_grp_biomed_hpc |
biomed_a100_gpu |
4 |
512 |
32 |
er_grp_biomed_hpc |
Standard node specifications
These are the standard specifications of our compute nodes available to all users. The gpu
partition contains
some additional A100s in nodes of other specifications.
Nodes |
CPU |
GPU |
Memory |
Partition |
33 |
128 |
- |
1024GB |
cpu |
10 |
36 |
4 x nvidia A100 40GB, w/ NVLink |
512GB |
gpu |