To apply, you must be a scientist in Swedish academia, at least at the level of PhD student.
Deadlines and Decisions
Proposals are processed weekly.
Note that staff will be on vacation during the summer and proposals submitted in July will be processed at a reduced pace.
This round is open for proposals until 2025-01-01 00:00.
|
Resource |
Centre |
Upper
Limit |
Available |
Unit |
Note |
|
|
Alvis |
C3SE |
1 000 |
80 000 |
GPU-h/month |
The Alvis resource is dedicated for AI/ML research.
|
|
|
The Alvis resource is dedicated for research in and research using AI/ML techniques.
For general use of GPU:s please use Dardel GPU instead, for generation of training data use Dardel or Tetralith.
The Alvis cluster is a national NAISS resource dedicated to Artificial Intelligence and
Machine Learning research.
Note: Significant generation of training data is expected to be done elsewhere.
The system is built around Graphical Processing Units
(GPUs) accelerator cards. The first phase of the resource has 160 NVIDIA T4, 44
V100, and 4 A100 GPUs. The second phase is based on 340 NVIDIA A40 and 336
A100 GPUs.
|
|
|
|
|
Tetralith |
NSC |
10 |
1 500 |
x 1000 core-h/month |
|
|
|
Projects will receive a default 500 GiB storage allocation on Centre Storage at NSC. If you need more storage, please apply for a Storage project and decline default storage from this compute proposal.
Tetralith is a general computational resource hosted by NSC at Linköping University.
Tetralith servers have two Intel Xeon Gold 6130 processors, providing 32 cores per server. 1844 of the servers are equipped with 96 GiB of primary memory and 64 servers with 384 GiB. All servers are interconnected with a 100 Gbit/s Intel Omni- Path network which is also used to connect the existing storage. Each server has a local SSD disk for ephemeral storage (approx. 200GiB per thin node, 900GiB per fat node). An IBM Spectrum Scale system comprises the centre storage. 170 of the Tetralith nodes are equipped with one NVIDIA Tesla T4 GPU each as well as a high- performance NVMe SSD scratch disk of 2TB.
|
|
|
|
|
Dardel |
PDC |
10 |
1 720 |
x 1000 core-h/month |
|
|
|
Dardel is a Cray EX system from Hewlett Packard Enterprise, based on AMD EPYC processors with an accompanying Lustre storage system.
The nodes are interconnected using Slingshot HPC Ethernet.
|
|
|
|
|
Dardel-GPU |
PDC |
200 |
6 160 |
GPU-h/month |
|
|
|
These GPUs are not nVIDIA GPUs but rather AMD GPUs, so if your software runs using CUDA, a certain amount of conversion of the code is needed.
You can read information about this at https://www.lumi-supercomputer.eu/preparing-codes-for-lumi-converting-cuda-applications-to-hip/
Reporting on GPU consumption on Dardel is not working yet.
Dardel-GPU is the accelerated partition based on AMD’s Instinct MI250X GPU of the Cray EX system from Hewlett Packard Enterprise. It has an accompanying Lustre storage system.
The nodes are interconnected using Slingshot HPC Ethernet.
|
|
|
|
|
Rackham |
UPPMAX |
10 |
1 500 |
x 1000 core-h/month |
Restrictive policy for NEW projects on Rackham.
|
|
|
Rackham will be decommissioned on 2024-12-31. No allocations will be made beyond this date. See https://www.uu.se/centrum/uppmax/nyheter/nyheter/2024-02-16-rackham-end-of-life
Few new Small-scale projects will be accepted. Continuation proposals are welcome. New projects must carefully describe a plan for the project from now until 2024-12-31. UU-affiliated projects can be moved automatically to the new local system. Other projects must have a clear and concrete exit plan.
Rackham provides 9720 cores in the form of 486 nodes with two 10-core Intel Xeon V4 CPUs each. 4 fat nodes have 1 TB of memory, 32 fat nodes have 256 GB, and the rest have 128 GB.
The interconnect is Infiniband.
|
|
|
