Skip to content
Platforms
From TU/e clusters to European supercomputers

No Umbrella maintenance scheduled Planned maintenance is announced here and in the News.

Platforms

Which platform fits depends on your workload, your data and who you are. Below is an overview of the platforms we support, from TU/e's own clusters to national and European supercomputers. Not sure where your work fits? Get in touch: we advise on the best platform, help with access and paperwork, and help you move up when your research grows.

  • Umbrella HPC Cluster TU/e

    The TU/e HPC cluster, built from departmental CPU and GPU resources and managed by the Supercomputing Center Infra Team.

    • For: TU/e employees and students with a TU/e account. Generic nodes are open to everyone at TU/e; departmental partitions to their own groups.
    • Best for: batch jobs with Slurm, simulations, data analysis, and interactive work through Open OnDemand.
    • Access: request an account.

    Umbrella documentation

  • Spike-1 TU/e

    NVIDIA DGX B200 platform for AI research: 4 nodes with 32 B200 GPUs, 250 TB DDN storage, Run:ai scheduling and hosted large language models.

    • For: TU/e researchers and students in the associated departments.
    • Best for: large-scale and distributed training, workloads that need a lot of GPU memory, and container-based AI workflows.
    • Access: application and intake with an SCC AI engineer.

    About Spike-1

  • Snellius National

    The Dutch national supercomputer, hosted by SURF.

    • For: researchers at Dutch institutions.
    • Best for: workloads that outgrow TU/e systems.
    • Access: through SURF. We can help you prepare your application.

    Snellius at SURF

  • EuroHPC European

    Pan-European supercomputers such as LUMI and Leonardo.

    • For: researchers and organisations in Europe.
    • Best for: the largest simulations and AI workloads.
    • Access: through EuroHPC access calls. We can help you prepare your application.

    EuroHPC

Already have access?