Student Cluster
We operate a student cluster equipped with GPUs used for teaching purposes, as well as for Bachelor's and Master's thesis projects. The cluster also includes CPU nodes for running standard CPU-based Jupyter sessions. More details about the hardware and cluster resources are in the technical information page.Access
Access to the cluster is restricted to:- Students taking courses that use the cluster, as well as the teaching assistants and mentors of such courses.
- Students doing Bachelor's and Master's thesis projects requiring GPUs, as well as their supervisors.
Time and Resource Restrictions
For each course or project a student will receive an individual time budget. By choosing a course tag or project tag for running a job or Jupyter notebook, the runtime is deduced from the respective budget. A course tag or project tag also comes with restrictions such as CPU count, RAM size, GPU model and count, job duration, etc. For large courses the time budget may also be applied to a team of users and restricting concurrent use.Use for Courses
Information for head TAs who wish to have their course to use the cluster is here. Access is granted per semester only. Access is revoked one week before the next semester starts and all data of the courses and the students taking them is removed.Use for Projects
Information for supervisors of Bachelor's and Master's thesis projects who wish to use the cluster for their students is here.Maintenance
We reserve the first week before a new semester starts for maintenance and upgrades. This is non-negotiable. The service unavailability must be taken into account of course or project planning.
- The operating system and its libraries and tools may be newer without backward compatibility.
- The default Cuda and cuDNN versions may be newer.