The table below shows the configuration of the compute nodes currently available on the system.
Node Type | Node Details |
---|
Compute Node (CN)
| 2 memory node types - Large memory nodes featuring 512 GB RAM
- Standard memory nodes with 256 GB RAM
|
Large memory nodes | Processor | - 2 x AMD 7742 per node (dual socket)
- 2.2 GHz at 225W (Turbo: up to 3.3 GHz)
- 64 cores per processor, 128 threads
|
Memory | - 512 GB RAM per node (~497 GB available for users)
- Cache:
- L1: 4 MB
- L1d: 2MB (64x32KB, 8-way set associative)
- L1i: 2MB (64x32KB, 8-way set associative)
- L2: 32 MB ( 64x512KB, 8-way set associative)
- L3: 256MB (16x16MB)
|
Network | |
Standard memory nodes | Processor | - 2 x AMD 7742 per node (dual socket)
- 2.2 GHz at 225W
- 64 cores per processor, 128 threads
|
Memory | - 256 GB RAM per node (~245 GB available for users)
- Cache:
- L1: 4 MB
- L1d: 2MB (64x32KB, 8-way set associative)
- L1i: 2MB (64x32KB, 8-way set associative)
- L2: 32 MB ( 64x512KB, 8-way set associative)
- L3: 256MB (16x16MB)
|
Network | |
Please have a look at the section Running Jobs of the Alps (Eiger) User Guide that shows how to allocate compute nodes on the system using the Slurm workload manager.