Skip to content

Hardware OverviewΒΆ

Important

Barbora NG documentation is a WIP. The documentation is still being developed (reflecting changes in technical specifications) and may be updated frequently.

The launch of Barbora NG is planned for October/November. In the meantime, the first computational resources have already been allocated in the latest Open Access Grant Competition.

Barbora NG consists of 141 non-accelerated compute nodes named cn[?-???]. Each node is a powerful x86-64 computer equipped with 192 cores (2x Intel Xeon 6952P with 96 CPU cores) and 768 GB RAM. User access to the Barbora NG cluster is provided by two login nodes login[1-2]. The nodes are interlinked through high speed InfiniBand NDR and Ethernet networks.

The parameters are summarized in the following tables:

In general
Architecture of compute nodes x86-64
Operating system Linux
Compute nodes
Total 141
Processor Type Intel Xeon 6952P
Architecture Granite Rapids
Processor cores 96
Processors per node 2
RAM 768 GB
Local disk drive no
Compute network InfiniBand HDR
non-accelerated 141, cn[001-141]
In total
Theoretical peak performance (Rpeak) ??? TFLOP/s
Cores 27072
RAM 108.288 TB