Pelle¶

Image of Pelle Svanslös, from https://www.comics.org/issue/105368/cover/4

Pelle is an upcoming general-purpose UPPMAX cluster, paid by Uppsala University.

Uppsala users of Rackham will be moved to Pelle by UPPMAX after applying to a Pelle project.

Non-Uppsala users of Rackham can move their data to Dardel, see the Rackham to Dardel migration guide.

Why use Pelle over another NAISS system?

Here is a table of why you want to use Pelle over another NAISS HPC cluster.

What is the status of Pelle?

Pelle is now open for all users for testing. Welcome!

Let us know if you are missing your favorite tools. We'll install those on demand.

What is different on Pelle in comparison to Rackham?

New operating system

Pelle is running Rocky Linux 9.
Most system software, including the Linux kernel, have newer versions compared to Rackham and other UPPMAX clusters.

Hardware

Pelle's compute nodes have much more CPU cores and memory compared to previous UPPMAX cluster.
All compute nodes have AMD Zen4 processors.
Details about the hardware is on the Pelle/Maja hardware page.

Software installations

Software available on Pelle is not included in the Software list on this web page yet.
Please use the module system to list available software, for example by running module avail or module spider.

Compute nodes

Info

You can reach all Rackham/Crex projects from Pelle!
Transition from Crex storage to Gorilla storage will be made from our side. You do not have to do anything!

How to apply to a Pelle project ¶

Pelle is quite similar to Rackham but we have prepared Pelle specific pages for some topics.

We are still preparing documentation of these subjects

As both clusters are UPPMAX clusters, we will transfer your data from Rackham to Pelle. Users will have to apply to a Pelle project.