Computing Resources
CDT in Data Science Computing Resources
The CDT has funding to support specialised computing or other equipment required for student research. In the first instance, students should consider whether the existing computing provision will suit their needs. If so, this saves the money and time required to order equipment.
Many of you will need to run jobs that are very compute and storage intensive. Learning how to run, say, 50 jobs in parallel is a significant advantage the week before a paper deadline. We have big computers to help you with this. Here is a partial list:
| James and Charles cluster | A dedicated computing cluster for CDT students |
| General purpose servers | There are general purpose servers in the School of Informatics. Some are for light tasks, like ssh and email. There are a few compute and big memory servers too, although they tend to be in heavy use. |
| Edinburgh Compute and Data Facility (ECDF) hosts a University-wide cluster with over 3000 cores. It can be a bit crowded and uses a slightly different Linux install than DICE. | |
| Supervisor Resources | Some supervisors have resources for their own students. Once you have a supervisor, ask them. Also, once you have a PhD supervisor, you will also be a member of an institute. Check whether your institute has any computational resources available. |
In the School of Informatics, we use AFS, which can be daunting. See the top AFS tips. In particular, see tip 7 if you need to use a long-running job that runs for more than 18 hours (this is not uncommon)!
If you would like the CDT to consider purchasing specialised computing or other equipment/resources required for your research, please complete a Funds request form and submit it as soon as possible.
If you have computing problems in Informatics, see the Computing help pages.