High Performance Computing
10 reasons to choose Bull HPC solutions
Products and solutions

NovaScale Clusters
Operating Systems
Cluster Management
Bull NovaScale   servers
Green HPC
Partners
Resource center

HPC competence
  centre
HPC newsletter
References
Brochures
Press

Bull NovaScale® Clusters

Bull NovaScale Clusters are pre-configured solutions tuned to deliver maximum performance while significantly reducing Total Cost of Ownership. Bull NovaScale Clusters also offer optimal flexibility and modularity, to match your precise requirements. They are available in a wide range of configuration sizes, from 3 to hundreds of nodes, and offer a broad choice of options for processors, interconnects, and operating environments.

Key features:

  • Optimal application performance and footprint: Bull NovaScale Clusters are tuned to deliver outstanding performance per watt and performance per square meter
  • Turn-key systems: Bull NovaScale Clusters are fully factory-integrated and -tested, and are easily and quickly integrated to your infrastructure
  • Modular, upgradeable solutions: additional modules can extend your configuration as your workload increases
  • Efficient cluster management: with the Bull NovaScale Master monitoring environment
  • High performance technologies: with the most recent Intel® Xeon® and Intel® Itanium® 2 processors

 

Bull NovaScale Cluster 400-W Series 400-W

Bull NovaScale Cluster 400-L Series 400-L

Bull NovaScale Cluster 3000 Series 3000

Cluster Nodes

Processor Family

Multi-core Intel® Xeon® Series 5100/5300

Multi-core Intel® Xeon® Series 5200/5400

Dual core Intel® Itanium® 2

Nodes

  • NovaScale R421/422 (compute nodes)
  • NovaScale R440/460 (service nodes)
  • NovaScale R421/422 E1 (compute nodes)
  • NovaScale R423 (service nodes)

NovaScale Intensive 3025/ 3045

Form Factor

1U-2U

1U-2U

4U

Memory

4GB up to 48GB

4GB up to 128GB

8GB up to 128GB

Internal Storage

2/3 SATA disks (R422/421)
3 SATA/SAS disks (R440)
6 SATA/SAS disks (R460)

2 fixed or 3 hot-swap SATA, or 3 hot-swap SAS disks (R421 E1)
2 hot-swap SATA disks (R422 E1)
6 SATA or 8 SAS disks (R423)

1 to 8 SAS hot-swap HDDs

Cluster

 

 

Packaging

19U racks

19U and 42U racks

Inter-connect

Infiniband - Gigabit Ethernet

Infiniband - Gigabit Ethernet

Configu-ration

3 packaged configurations from 2 to 16 nodes

Large choice of configurations up to hundreds of nodes

Operating Software

Windows CCS 2003 (Windows HPC Server 2008 coming soon)

Red Hat Linux

Development tools

Visual Studio

Intel Processor Family compilers: C/C++, Fortran and Math. Kernel and Integrated/ Performance Primitive librairies, GNU Compilers: C, Fortran 77

Platform and System Management

Bull NovaScale Master System Management
Microsoft Compute Cluster Pack

Bull Advanced Server HPC Edition & Bull NovaScale Master System Management

Bull NovaScale Cluster 400-W    Back to top

 

Access the world of High Performance Computing with NovaScale Clusters and Windows CCS 2003 (Windows HPC Server 2008 coming soon)

Your organization is already running a Microsoft environment, your engineers are familiar with Windows? You can rely on Bull’s expertise to move to HPC with Windows Compute Cluster Server 2003, and boost your compute capacity while maximizing the use of your existing infrastructure.

 

Bull NovaScale Cluster 400-W combines:

  • The performance and density of Bull NovaScale R4xx servers equipped with Intel® Xeon® processors
  • The advantages of Windows CCS 2003
  • Pre-installed Operating System and Licenses
  • Expert support from Bull’s Microsoft Competence Center

A broad choice of applications are already supported and certified for Windows CCS, such as structure analysis, computational fluid dynamics, and crash test software.

  • Bull NovaScale Cluster 400-W4
    An affordable and upgradeable solution for a first investment in High Performance Computing, with 2 to 4 compute nodes that will relieve the strain on your workstations
  • Bull NovaScale Cluster 400-W8
    An attractively-priced and easily-maintained solution that will give independent compute resources to a small team of users. 4 to 8 compute nodes will handle large jobs or several jobs simultaneously.
  • Bull NovaScale Cluster 400-W16
    The ideal solution to equip a workgroup with independent high performance computing resources. 8 to 16 compute nodes will handle the compute workload of all users more than adequately.

Bull NovaScale Cluster 400-L    Back to top

 

The Bull NovaScale Cluster 400-L pre-configured solutions combine:

  • The performance and density of Bull NovaScale R42x servers equipped with Intel® Xeon® processors
  • The advantages of Bull’s  integrated, optimized Linux® environment
  • And expert support from Bull’s HPC Competence Center
 

  • Bull NovaScale Cluster 400-L24
    3 to 24 nodes - “Workgroup”
    A compact and affordable solution to equip a workgroup with independent high performance computing resources
     
  • Bull NovaScale Cluster 400-L64
    24 to 64 nodes - “Departmental”
    The ideal solution to enable a department to handle large jobs or several jobs simultaneously
     
  • Bull NovaScale Cluster 400-L96
    64 to 96 nodes - “Enterprise”
    A very powerful solution for a large Computing Center
     
  • Bull NovaScale Cluster 400-L+
    Beyond 96 nodes - “Capability”
    Outstanding performance to match outstanding requirements, with a modular packaging that allows the configuration to grow as the workload increases
     

Bull NovaScale Cluster 3000    Back to top

 

The Bull NovaScale Cluster 3000 Series based on NovaScale 3xx5 servers offers:

  • The outstanding performance of Intel® Itanium® 2 processors for the most complex instructions and for all applications that benefit from a large shared memory
  • The advantages of Bull’s integrated, optimized Linux® environment
  • And expert support from Bull’s HPC Competence Center
 

  • Bull NovaScale Cluster 3000-L64
    3 to 64 nodes - “Departmental”
    The ideal solution to enable a department to handle large jobs or several jobs simultaneously
     
  • Bull NovaScale Cluster 3000-L96
    64 to 96 nodes - “Enterprise”
    A very powerful solution for a large Computing Center
     
  • Bull NovaScale Cluster 3000 L+
    Beyond 96 nodes - “Capability”
    Outstanding performance to match outstanding requirements, with a modular packaging that allows the configuration to grow as the workload increases

 

 
LEARN MORE
 

Intel Cluster Ready: guaranteed application mobility
Bull NovaScale Cluster running under Bull Advanced Server 5 certified by Intel

Intel flyer "Bull joins Intel Cluster Ready Program"




NEED INFORMATION?
  Contact us.
NEWS

Bull and the CEA sign a collaboration contract for the design and acquisition of a petaflops-scale supercomputer

 
SUCCESS

Bull co-ordinates a major project in the fight against cancer

GENCI and CEA order a Bull supercomputer to speed up innovation and fundamental research

Making the impossible possible – University of Cardiff’s new High Performance Computer

 
LEARN MORE

New technical white paper co-authored by Bull and Microsoft HPC experts: A Hybrid OS Cluster Solution - Dual-Boot and Virtualization with Windows HPC Server 2008 and Linux Bull Advanced Server for Xeon

 

 

Contact  |  Site map  |  Legal  |  Privacy