# Compute Ecosystem

P1 funds and helps run partitions at partner facilities, and we point the way to the rest. The Danish compute
landscape is bigger than P1. As a consortium we know it well: we understand the challenges of
getting compute for research, we have experience across the systems below, and when your needs fit
better elsewhere we point you in the right direction.

**We consult and guide. We don't broker access.** Getting onto the systems below goes through
their own institutions and application processes, but if you are unsure where your project
belongs, [talk to us](/about/) first.

_(Diagram: a researcher enters through P1, which consults and guides, and points on to university, national, and international HPCs.)_

## University HPCs

Most Danish universities operate their own HPC facilities, generally available to their own
researchers and collaborators:

- **UCPH**: SCIENCE HPC Centre, Hendrix
- **DTU**: DCC/G-Bar Central, Niflheim, Sofia, Steno, plus 5 department systems
- **SDU**: [UCloud](https://cloud.sdu.dk/), Bitten
- **Aarhus**: [GenomeDK](https://genome.au.dk/)

## The Danish Research Network

Danish universities and national facilities share a dedicated high-bandwidth network,
forskningsnettet, operated by [DeiC](https://www.deic.dk/en/danish-research-network). It is why
moving a dataset between two Danish facilities is usually faster than pulling it down to your own
machine and pushing it back up, and why a cluster-to-cluster hop is worth preferring when you have
the choice.

_(Diagram: forskningsnettet drawn as one high-bandwidth link, with the DTU, NGC, and Gefion facilities hosting the P1 partitions above it and the UCPH, SDU, Aarhus, and Computerome facilities below it.)_

The diagram shows only the systems documented on this site. Forskningsnettet reaches considerably
further, covering the other Danish universities and research institutions.

## National HPCs

- [DeiC](https://www.deic.dk/): Danish e-Infrastructure Consortium, coordinating national HPC
  resources and granting compute allocations to researchers at Danish universities
- [Computerome](https://computerome.dk/): the Danish national supercomputer for life sciences,
  with support for sensitive data

## International HPCs

- [LUMI](https://www.lumi-supercomputer.eu/) (Finland, EU): one of Europe's largest
  supercomputers, accessible to Danish researchers through DeiC allocations

## Where P1 fits

The P1 partitions are the entry point for P1-affiliated researchers: start on the
[DTU HPC](/clusters/dtu/) for general work, the [NGC HPC](/clusters/ngc/) for sensitive data, and
the [Gefion sandbox](/clusters/gefion/) to get a workload ready for large-scale runs. When your
needs outgrow them, the systems above are the next step, and we are happy to help you choose.