Mimir
Introduction
The Mimir cluster consists of 10 compute nodes divided into two partitions. The partitions are mimir and himem-mimir, where the mimir partition consists of a total of 9 compute nodes and the himem-mimir partition consists of 1. The Mimir cluster is generally used for research related to the life sciences that do not require a lot of compute power but instead require a lot of RAM usage. If that matches your requirements, you can send a formal application to irei@hi.is.
More information on the formal application can be found in the Etiquette chapter.
The himem-mimir stands for HIgh MEMory mimir partition. This partition provides a compute node that can compute jobs that require high memory usage.
Hardware Specification
In total, the Mimir partition has 256 cores and 960 GB of memory available. 4 x standard compute nodes and 1 x GPU node:
| Count | Name | Cores/Node | Memory/Node (GB) | Features |
|---|---|---|---|---|
| 9 | mimir | 64 (2x32) | 256 (252) | Intel(R) Xeon(R) Platinum 8358 |
| 1 | mimir-himem | 64 (2x32) | 2048 (2044) | Intel(R) Xeon(R) Platinum 8358 |
More information can be found in the specs chapter and the partitions chapter.
Request Access
You can find all information on requesting access to the Mimir partition in the Etiquette chapter.
Recommended chapters
Here are recommended chapters to help you get started on using Mimir
-
Transfer Data to the Mimir Disk: read here
-
Submitting Slurm Jobs General usage on submitting a job
-
Connection via SSH: Read about SSH keys