2016822554

 

The DLTM high-performance computing (HPC) centre offers resources and services that solve complex problems using parallel computation, in basic and applied research in areas involving high-tech systems simulation and design.

It provides tools and resources to the following research and manufacturing sectors:

  • Fluid dynamics/hydrodynamics
  • Structural computing
  • Design and simulation
  • Digital modelling
  • 3D modelling and rendering
  • Big data

 

Its activities focus on:

  • Parallel computing support;
  • Consultancy service for HPC architecture design and/or management;
  • Processing, management and analysis of Emso station data (sea lab);
  • Big data, support for the use of SkyLable object open source storage.

 

COMPUTING CENTRE

The supercomputing centre is the essential infrastructure resource on which the centre's three laboratories are based (ship lab, sea lab, HPC lab).

The district has a Dell cluster, purchased in July 2016, consisting of a hybrid HPC and cloud architecture, in turn made up of the following:

  • HPC Cluster
    • Hw:
      • 18 computing nodes (Dual CPU - Intel 52695 V3 2.4 Ghz) each with 28 cores and 64 GB of RAM, giving a total of 504 cores and more than 1 TB of RAM
      • Mellanox INFINIBAND 56 Gb/s FDR high-performance connection infrastructure
      • 1 Head Node with 24 cores and 386 GB of RAM
      • 2 NVIDA K40M cards
      • High-performance shared storage, 21 TB of net space.
      • Computing power (excluding GPU): more than 17 teraflops
    • System SW:
      • Bright Cluster Manager
      • CentOS 7
      • PBS Pro
      • Principal libraries and compilers
  • Private Cloud
    • Hw 4 nodes (Dell FC630):
      • Dual CPU Intel E52650 v3 with 128 GB of RAM
      • For a total of: 80 cores and 512 GB of RAM
    • Hw 4 nodes (Dell FC830):
      • Four CPU Intel E54620 v3 with 256 GB of RAM
      • giving a total of 160 cores and 1 TB of RAM
      • The cloud resources pool has a total of 240 physical cores and 1.5 TB of RAM
      • High-performance storage based on MS storage spaces in Scale Out File Server with 55 TB and 2.9 TB SSD for the cache.
    • System SW:
      • System Center 2012 (soon to be upgraded to 2016)
      • Azure Pack portal
      • Number of unlimited licenses of Windows Server 2012 and 2016

 

Dell     Microsoft System Center     Microsoft Azure Stack

 

SERVICES

High-reliability HPC resources for a high level of processing power (CPU and memory intensive), sold in pay-per-use mode, based on a list available from This email address is being protected from spambots. You need JavaScript enabled to view it.

Members of DLTM, research bodies and non-profits receive a special discount on the list price.

The following CFD and FEM simulation software is available in SaaS mode on the HPC cluster, including for commercial use:

  • Ansys CFD (Ansys CFX, Fluent) and Ansys Mechanical
  • Starccm+
  • MSC suite (Nastran, Patran, Marc, Apex, etc...)
  • OpenFOAM

 

On request, we can install any other software that uses parallel architecture.

The computing will be performed in batch mode, and, on request, the user will be able to use a virtual graphics workstation for the calculation and/or realisation of the Mesh.

 

ANSYS   SIEMENS CD-adapco   MSC Software   OpenFOAM - The OpenFOAM Foundation

 

High-reliability, high-performance cloud resources are offered by the Microsoft infrastructure, which delivers IaaS services using virtual machines to meet user needs, both with Windows Server 2016 (with unlimited licences) and Linux for any type of software and service sold to third parties, for example:

  • Web server
  • Database
  • Gestionalewindows--2016 server
  • Machine learning
  • Big data

 Windows Server 2016

For more information, write to the laboratory coordinator: This email address is being protected from spambots. You need JavaScript enabled to view it.

 

 

torna all'inizio del contenuto