EOS is the new high-performance computer cluster (HPC) of the University of Pavia, installed at the premises of the UniPV Technological Infrastructure Management Service.

It consists of a multiprocessor and multi GPU cluster with 160TB parallel storage system and with a computational total of 672 Intel CPU Cores, 71,680 Nvidia Cuda Cores, 8,960 Nvidia Tensor Cores, divided into:

  • 7 FAT nodes (each with 768 GB RAM);
  • 7 GPU nodes (each with 128 GB RAM and 2 Nvidia V100 GPUs with 32GB RAM);
  • 7 standard nodes (each with 128 GB RAM).

The nodes are connected by an Infiniband network and Linux Operating System.
It is available to the scientific community of the University and meets the needs of high-performance computing, parallel computing, big data and data science. GPU nodes are of particular interest for mathematical modeling projects, numerical simulations, machine learning applications, and artificial intelligence.

HPC guide
Introduction to Unix
Cluster access request form

Financing

EOS was mainly funded by the Excellence Project of the Mathematics Department and the UniPV PASS fund, with further contributions from six Departments of the University and the IMATI – CNR Institute.