Welcome to HPC2N’s Support and Documentation pages¶

These pages are for user support and documentation, including documentation and tutorials for the software installed on HPC2N’s cluster, Kebnekaise.

Please see the main HPC2N website for general information about HPC2N (High Performance Computing Center North), as well as contact information.

Current system

Kebnekaise, the main HPC2N computing resource deployed during 2016/2017 and extended in 2018 and 2023, is a heterogeneous system with Skylake CPUs, Skylake V100 GPU nodes, 3TB memory large memory nodes, AMD Zen3 nodes, and A100 GPU nodes. The V100 and A100 GPUs both offer CUDA cores and Tensor cores. For a scalable parallel performance, the system is equipped with high bandwidth, low latency EDR/FDR InfiniBand interconnect.

Kebnekaise is now a local compute system, accesible only for researchers in groups with a PI that is affiliated with UmU or HPC2N’s partner sites (IRF, SLU, MIUN, LTU).

The compute cluster will be extended with new hardware gradually.

If you want to read more about Kebnekaise’s architecture, go directly to the Kebnekaise hardware page on the HPC2N main pages.