XHG Nodes¶
We have 6 Dell PowerEdge XE7745 nodes with Nvidia Hopper H200 141GB cards for GPU jobs.
Accessing the XHG nodes
5 out of 6 XHG nodes are allocated to the Apini cluster, purchased by Digital Environment Research Institute. The other XHG node is restricted to users within the Science and Engineering (S&E) faculty only.
GPU high memory
The XHG nodes have more RAM available than standard GPU nodes, supporting
up to 320G RAM per GPU (16 tasks per GPU). Authorised users may use the
--mem-per-cpu=20G submission parameter to request larger amounts of RAM
on these nodes.
| XHG | Dell PowerEdge XE7745 |
|---|---|
| Processor | 2 x 64 Core AMD Zen EPYC 9555 |
| Cores/Node | 128 |
| RAM | 1.5TB |
| Accessible RAM | 1.4TB |
| TMP Size | 2.7TB |
| Interconnect | 100Gb Ethernet |
| GPU | 4 x NVIDIA Hopper H200 |
| GPU architecture | Hopper |
| Form Factor | NVL (PCIe) |
| Tensor Cores | 528 4th Generation |
| CUDA Cores | 16,896 |
| GPU Memory | 141GiB per GPU |
| CUDA Compute | 9.0 (CUDA version 11.8 or greater required) |
