• Le cloud by Bull
• |
• The structured cloud
• |
• The agile cloud
• |
• The secure cloud
• |
• The controlled cloud
• |

• Bio Data Center
• |
• Infrastructure Services
• |
• Servers
• |
• Storage
• |
• Networks

• Virtual Shore
• |
• Consulting
• |
• Systems integration
• |
• BI
• |
• ERP
• |
• Open Source

• Agile Security
• |
• Amesys
• |
• Security Consulting & Integration
• |
• HA
• |
• Network Security
• |
• IAM
• |
• PKI
• |
• Payment Security

• Outsourcing services
• |
• Insourcing
• |
• Software support
• |
• Agarik

A new generation of SMP systems
Building on its acknowledged experience in designing SMP systems, Bull is now extending its bullx range of Extreme Computing nodes. With the bullx supernode Bull introduces a new generation of SMP system that can satisfy the needs of even the most memory-hungry applications.

Whether you need an efficient SMP node within your High Performance Computing cluster, or you are looking for versatile large nodes to simplify the architecture of a large-scale cluster, the bullx supernode will deliver the ultra-high performance your applications need. The bullx supernode family was designed and developed by Bull’s Extreme Computing teams with the following guidelines in mind: Optimization, compactness and simplification of the compute node for HPC usage Optimization and enhanced connectivity of the service nodes Versatility, to support a wide range of applications, including memory-hungry code Expandability: up to 4 x 4-socket SMP nodes can be interconnected through a Bull-designed switch, to form a large SMP node - with up to 16 sockets and 4 TB of memory

The bullx supernodes capitalize on the latest technological advances such as:

• New-generation Intel® Xeon® E7 Family processors (code name: Westmere-EX), designed for high-end servers, with advanced reliability and exceptional scalability
• Intel Quick Path Interconnect protocol
• Optimized intra-node throughput and latency thanks to the bullx MPI
• InfiniBand QDR network connection
• Support for GPU systems

Thanks to these innovations, the bullx supernode delivers:

• Improved performance (enhanced compute node efficiency, reduced latency, higher communication and I/O throughput)
• Lower cost of ownership (fewer components, lower management cost, reduced installation time, power efficiency)
• Enhanced reliability (“fat” nodes mean fewer nodes)

Power-conscious design:

• Processor power management features
  
• Sleep mode
  
• Ultracapacitor protecting the node from power outages up to 300ms. In areas with good quality power supplies, the ultracapacitor means a UPS is no longer needed – a gain of up to 15% on power consumption!
  
• High efficiency power supply unit

High-density compute nodes

The exclusive format of the bullx S6010 nodes – 1.5 U L-shaped drawers that fit together in pairs to form a 3U drawer – delivers outstanding density for a high-end HPC node. They are the ideal platform to build large-scale HPC clusters while minimizing the number of nodes, and thus simplifying the infrastructure and operation of the whole system. More information on bullx S6010

Service nodes: enhanced I/O and storage capabilities within 3U

The bullx S6030 nodes offer advanced connectivity features, an enhanced power supply, extended storage options and redundancy features. So they are ideally suited to act as management or I/O nodes. They can also be coupled to a GPU system. More information on bullx S6030