Apply to this round for a project on the Berzelius SuperPOD at NSC. Berzelius is a non SNIC system donated to NSC by Knut and Alice Wallenberg foundation (KAW). Monthly evaluation of proposals during the year.
To apply, you must be a scientist in Swedish academia, at the level of PhD student or higher.
Applications are normally evaluated during the last week each month.
Submit your proposal at least one week before the end of a month to be considered for an allocation from the first of the following month. Received proposals will be evaluated against each other and time that become available as project ends at the end of a month will be allocated to the proposed projects accordingly.
Berzelius is an NVIDIA SuperPOD consisting of 60 DGX-A100 node, sporting a total of 480 NVIDIA A100 GPUs.
The SuperPOD uses the SLURM resource manager and job scheduler. All DGX-A100 nodes have 8x NVIDIA A100 GPUs (40GB), 128 CPU cores (2x AMD Epyc 7742), 1 TB of RAM and 15 TB of NVMe SSD local disk. High performance central storage is available using 4x AI400X DDN, serving 1 PB of storage space to all nodes of the cluster. All DGX-A100 GPUs have dedicated Mellanox HDR InfiniBand HBAs, that is, there are 8 Mellanox HDR HBAs per DGX-A100 node, connected in a full bisection bandwidth, fat-tree topology.
Applications are normally evaluated during the last week each month.
Submit your proposal at least one week before the end of a month to be considered for an allocation from the first of the following month. Received proposals will be evaluated against each other and time that become available as project ends at the end of a month will be allocated to the proposed projects accordingly.
High performance central storage is available using 4x AI400X DDN, serving a total of 1 PB of storage to all nodes of the cluster via a dedicated InfiniBand interconnect. Aggregate IO performance is 192 GB/s from the central storage and the dedicated data interconnect bandwidth per node is 25 GB/s.
NSC centre storage (as available on Tetralith) is not accessible on Berzelius.
Click above to show more information about the resource.
