Home > News > Members and Partners News > CLUSTERVISION

CLUSTERVISION News

December 7, 2015 - ClusterVision is growing

Present in France for about 10 years, ClusterVision is proud to announce the inauguration of its new office in Bordeaux. From there, the French team will expand its coverage of projects in France, Southern Europe and beyond. Developing local presence being a top priority, this new office will allow ClusterVision to settle durably within the French HPC landscape.

New cluster installations :

  • Magdeburg University

The University of Magdeburg has placed an order for a 3000 cores Intel HasWell Cluster with ClusterVision. Connected with an Infiniband Intel TrueScale network, it is also equipped with a 320TB BeeGFS storage filesystem.  In order to deploy and manage this Cluster, the University selected Trinity, the open source solution developed by ClusterVision.

  • Volvo IT

After several successful installations, Volvo IT and ClusterVision keep working together around the installation of a new cluster. This time, more than 700 cores were provided to the Gothenburg research centre to bring its scientists new calculation perspectives.

  • Max Planck Institute for Gravitational Physics (Albert Einstein Institute)

The biggest research institute dedicated to Relativity, the Max Planck Institute has placed an order with ClusterVision to install a new 9500 cores cluster divided over 600 nodes. This cluster will be connected through Intel’s Omni-Path network and will be equipped with a 500TB BeeGFS storage filesystem.
The Cluster, located in Potsdam, will empower the various scientific teams with a tool which matches their needs and their ambitions.

The cloud brings potential game changing possibilities to HPC which is why Clustervision developed Trinity, a new open source management software based on OpenStack.

Trinity brings the flexibility of the cloud to your HPC cluster without any loss in performance: You can now allocate hardware resources to different projects o users and manage your virtual clusters in an independent and fully secured way.

Features

Power of HPC

  • Scalable to tens of thousands of nodes
  • Full hardware support (IPMI, Infiniband, PXE)
  • Complete HPC stack (job schedulers, MPI & scientific libraries, compilers, profilers, debuggers)
  • Scheduler - SLURM
  • MPI libraries - OpenMPI, Intel MPI, MVAPICH, MPICH2
  • Compilers - GNU and Intel
  • No performance loss (Docker based containerization)
  • Load balancing (HPC) partitions
 
Agility of the Cloud
  • Full control of environment customization for users with different requirements
  • Ease of management and usage metering (Trinity dashboard)
  • Allows customers to host their own private or public IaaS cloud (for general IT, selling cycles, or third party collaborations)
  • Ability to rapidly re-partition the cluster according to current internal demands

 

© Teratec - All rights reserved - Lawful mention