MARCA  logo usm

High Performance

High Performance

labs header hpcd

lab2 480

lab2 320

labs equipamiento

High Performance Computing Data Center

The Computing and IT areas count with a data center that it´s home to CCTVal´s computing cluster. The main purposes of this data center are:

  • To execute specialized programs in Physics and Engineering (data analysis, simulation of particle reactions and more).
  • To develop new applications in High Performance Computing.
  • To serve as a calculation node for experiments conducted in CERN (LHC and ATLAS).
  • To provide network infrastructure (DND, Web, proxy, etc.) and services of cluster operations (security, batch system, grid services, etc.).

HPC Data Center counts with the following equipment

HPC CPU: 24 nodes of work (448 CPU cores):

  • 16 nodes x 16 CPU cores 2.80 GHz
  • 8 nodes x 24 CPU cores 3.06 GHz

HPC GPU: 4 nodes of work (80 CPU cores, 2752 GPU cores):

  • 2 nodes x 16 CPU cores 2.27 GHz, 480 GPU cores 0.80 GHz (2 x nVidia Tesla C1060)
  • 2 nodes x 24 CPU cores 2.67 GHz, 896 CPU cores 1.15 GHz (2 x nVidia Tesla M2050)

HPC servers of services (80 CPU cores):

  • 10 nodes x 4 CPU cores 1.60 GHz
  • 2 nodes x 2 CPU cores 2.66 GHz
  • 2 nodes x 4 CPU cores 2.33 GHz
  • 6 nodes x 4 CPU cores 3.07 GHz
  • 1 node x 4 CPU cores 3.30 GHz

HPC storage:

Grid Storage (200 TB):

  • head     20 TB (30 TB raw disk)
  • pool 1   60 TB (72 TB raw disk)
  • pool 2   60 TB (72 TB raw disk)
  • pool 3   60 TB (72 TB raw disk)

Local Storage (54 TB):

  • pool 1  40 TB (48 TB raw disk)
  • pool 2   8 TB (12 TB raw disk)
  • pool 3   8 TB (12 TB raw disk)