The Teaching cluster is used to support teaching courses, especially MSc courses and the Machine Learning Practical. It has now been merged into the Informatics Compute Facility (ICF) as the Teaching partition.
The Teaching partition has 155 GPUs in 13 servers. This includes 8x A40 48GB GPUs (crannog01-02), 8x NVIDIA RTX A6000 40GB GPUs (landonia11), 48x NVIDIA RTX 2080 Ti 11GB (damnii07-12), and 8x H200 141GB partitioned into smaller virtual GPUs (saxa). Please note that the configuration may change. Check the MOTD and the blog for details.
The status and resources of the ICF, using Slurm-Web, can be viewed at https://icfwebview.inf.ed.ac.uk
Resources in a partition can be seen by clicking on Resources in the left menu bar then filtering by partition. Here the GPUs and cores that are in use on the compute nodes can be viewed by expanding the diagram of the nodes.
How to get access
Undergraduate and taught MSC Students will get access based on the courses they are on. PGR students and staff will get low priority access when the nodes are not being used for a taught class.
To use the cluster for a project, please ask your supervisor to create a Computing Support ticket giving details of the project.
How to use the cluster
Use of this cluster is controlled by Slurm. First ssh to the head node (icf or icf2) then use Slurm commands.
Connections to the cluster are restricted by the informatics firewall. You can connect to the head nodes from the SSH gateways, other servers inside the Informatics firewall, DICE desktops and when using the School's OpenVPN..
Here's how to use a cluster without breaking it:
When submitting jobs with srun if a partition is required you should use the -p Teaching option.
Files and backups
- Jobs on the compute nodes cannot access your files in AFS. Instead there are home directories on a local distributed filesystem. You can copy files from AFS by logging in to a head node then using the cp command. AFS home directories can be found under
/afs/inf.ed.ac.uk/user/. - There is no disk quota but space is limited so please delete files you have finished with.
- Your files are NOT backed up. DO NOT USE THIS FILESYSTEM AS THE ONLY STORAGE FOR IMPORTANT DATA. Disk space will be reclaimed once your access has finished. It is your responsibility to make copies of anything you wish to keep. There is some redundancy for disaster recovery.
- The current distributed filesystem is Lustre (wikipedia). Some users are still using the previous GlusterFS (wikipedia) filesystem.
Software
The cluster runs the Ubuntu (24.04 Noble) version of DICE. If you would like software to be added, please ask Computing Support.