PLX and Galileo: updates

Dear Users,

as recently announced, the new Tier-1 CINECA cluster Galileo will be soon made available to users, and will replace PLX that is due off-line on January 23^rd . All the projects and usernames still active on PLX  by this date will be moved automatically to the new cluster with the remaining budget. All PLX data will be made available on Galileo.

Storage areas on Galileo will have $HOME, $CINECA_SCRATCH and $WORK as usual. More details on quotas and disks space available will be found on the login message of the new system and on updated on-line guides.

The new cluster will be on line on January 26^th for a pre-production period and will be actually available for production runs on February 2^nd . Galileo will have an architecture similar to PLX and will be equipped with 516 compute nodes ( 8256 cores) each with 2 x Intel Xeon 2630 v3 @ 2.5 GHz, 8 cores each (16 core/node), with 128GB/node RAM DDR3 1866MHz. 384 nodes will also host 2 Intel Phi 712p cards. Total storage will be 2.5 PB.

A detailed guide of the new cluster will be available our HPC portal by January 19^th .

Regards,

HPC User Support @ CINECA