Logo Cineca Logo SCAI

You are here

GALILEO

 

The GALILEO supercomputer has been introduced first in January 2015 and it has been available to Italian public and industrial researchers until January 2018. It has been the the national Tier-1 system for scientific research.
Starting from January 2018 GALILEO has been reconfigured with Intel Xeon E5-2697 v4 (Broadwell) nodes, inherited from MARCONI system.
Starting from March 2018 GALILEO is again in production, available for italian research community.

System Architecture

Compute Nodes
:

  • 1022 36-core compute nodes. Each one contains 2 18-cores Intel Xeon E5-2697 v4 (Broadwell) at 2.30 GHz
  • 60 compute nodes equipped with 1 nVidia K80 GPU
  • 2 compute nodes equipped with 1 nVidia V100 GPU

All the compute nodes have 128 GB of memory.

Login and Service nodes: 8  Login node are available. 8  service nodes for I/O and cluster management.

All the nodes are interconnected through a Intel OmniPath with OPA v10.6 capable of a maximum bandwidth of 100Gbit/s between each pair of nodes.

 

GALILEO

UserGuide

Model: IBM NeXtScale cluster
Architecture: Linux Infiniband cluster
Network: Intel OmniPath (100Gb/s) high-performance network

Nodes: 1022 (Intel Broadwell)
Processors: 2 x 18-cores Intel Xeon E5-2697 v4 at 2.30 GHz
Cores: 36 cores/node, 26.572 cores in total
RAM: 128 GB/node
+
Nodes: 60 (Intel Broadwell) with GPUs
Processors:
2 x 18-cores Intel Xeon E5-2697 v4 at 2.30 GHz + 2 nVidia K80 GPUs
Cores: 36 cores/node, 2160
cores in total
RAM:
128 GB/node
+
Nodes: 2 (Intel Broadwell) with GPUs
Processors:
2 x 18-cores Intel Xeon E5-2697 v4 at 2.30 GHz + 2 nVidia V100 GPUs
Cores: 36 cores/node, 72
cores in total
RAM:
128 GB/node