CLAIX-2025 Compute Nodes
On this page, you will find information about the available compute nodes for the CLAIX-2025 cluster.
CLAIX-2025
The CLAIX-2025 cluster is the current computing component of the HPC system CLAIX (wird in neuem Tab geöffnet). The commissioning of CLAIX‑2025 was successfully completed in June 2026.
The cluster is funded through two sources: Tier-2 NHR (wird in neuem Tab geöffnet) and WestAI (wird in neuem Tab geöffnet). Hence it is accessible to NHR- and WestAI computing time projects only.
It consists of two segments.
- the traditional High Performance Computing (HPC) Segment, equipped with AMD Genoa 9654 (wird in neuem Tab geöffnet) CPUs,
- the Machine Learning (ML) Segment, equipped with AMD Turin 9535 (wird in neuem Tab geöffnet) CPUs and additional NVIDIA H100 (wird in neuem Tab geöffnet) GPUs for accelerated ML workloads. The ML segment can be used for traditional HPC workloads as well, of course.
The exact specifications are displayed in the table below.
The HPC Segment offers compute nodes in two memory configurations. All nodes have local SSD storage, that can be used with BeeOND. You can select a node that meets your requirements by specifying the appropriate partition in your batch job.
Hardware Specification
| Segment | Per-node hardware specs | Nodes | TFLOPS | Available resources / year | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| HPC |
| 421 | TBD | NHR 708 Mio. Core-Hours p.a. | ||||||||||
| ML |
| 20 | TBD | NHR 210 k GPU-Hours p.a. WestAI 491 k GPU-Hours p.a. |
