MARCONI100: updates on the new accelerated GPU cluster

Dear Users,
This is to inform you that the installation of the new accelerated partition of CINECA (Marconi100) proceeds as scheduled. The new racks have been delivered to CINECA and will be installed soon. We remind you that the old MARCONI-A2 (KNL) partition will be out of production within the end of January 2020 and Marconi100 will start its production phase in March 2020.

MARCONI100, as announced, is based on the IBM Power9 architecture with NVIDIA Volta GPUs. Specifically, each node will host  2×16 cores IBM POWER9 AC922 at 3.1 GHz with 256 GB/node of RAM memory and 4 x NVIDIA Volta V100 GPUs per node, Nvlink 2.0, 16GB. The number of nodes will be 980, totallying 31360 cores. Internal Network: Mellanox Infiniband EDR DragonFly+
To help users porting their own code on this new GPU accelerated architecture, the 2020 updated catalogue of available training courses, including CUDA programming, is available and can be consulted here
Best Regards,
HPC User Support @ CINECA