NASA's 10,240 CPU Supercomputer Takes Off with Voltaire
Voltaire Grid Backbone solutions power NASA's Columbia supercomputer, the world's largest "production" supercomputer. Holding the number-four position on the Top500 list of the world's fastest supercomputers, the system achieved sustained performance of 51.87 trillion calculations per second (teraflops) and an 88 percent efficiency rating on the LINPACK benchmark.

The Columbia supercomputer is expected to revolutionize the rate of scientific discovery at NASA by reducing the computation time required to solve some of the world's most complex scientific problems.

Columbia was deployed in fewer than 120 days, of which, only three weeks were needed to cable and bring up the InfiniBand fabric.

• Voltaire Grid Director ISR 9288, a 288 port 4X (10 Gbps) multi-service switch
  
• GridVision InfiniBand Fabric Management Software
  
• 20 SGI® Altix® 512 processor systems

"NASA, Intel, SGI, and Voltaire have built the fastest, most efficient cluster in the world - and in only 120 days. The scalability and ease of use of Voltaire's InfiniBand interconnect solution and their dedicated support through the deployment process were key factors in achieving world class performance and efficiency for the Columbia supercomputer."

Walt Brooks
Division Chief, Advanced Supercomputing Division, NASA

Click here to download a full case study on NASA (544 K)



Tokyo Institute of Technology Picks Voltaire for Supercomputer
Japan's Tokyo Institute of Technology (Tokyo Tech) selected Voltaire's InfiniBand-based Grid Backbone™ switching solutions as the high-performance interconnect for a powerful Sun Microsystems supercomputer that is used for computational scientific research.

The Tokyo Tech system is Japan's largest grid and is the seventh largest supercomputer in the world as ranked by the Top500.

The system uses next generation Sun Fire™ x64 (x86, 64-bit) servers using over 10,000 AMD Opteron processor cores and connected by multiple Voltaire Grid Director™ ISR 9288 switches. The Voltaire switches also deliver high-performance connectivity to more than 1 Petabyte of Sun and NEC storage.

"We expect to achieve leading-edge performance and scalability for this clustered supercomputer with the combination of multi-processor Sun Galaxy platforms and Voltaire's multi-service Grid Backbone switches that offer proven InfiniBand performance. This supercomputer will enable us to take on larger and more complex computing problems that will advance scientific research in areas ranging from engineering to weather to medicine."

Professor Satoshi Matsuoka
Global Scientific Information and Computing (GSIC) Center, Tokyo Tech



Lawrence Livermore Selects Voltaire DDR InfiniBand Solutions for Peloton Supercomputer
Lawrence Livermore National Laboratories (LLNL) selected Voltaire Grid Backbone solutions as part of an Appro clustered supercomputer for the Peloton Linux Cluster Project. The 16,128 processor supercomputer leverages Voltaire's multi-service director-class switches with double data rate (DDR), 20 gigabit per second InfiniBand to provide industry-leading performance for LLNL's complex computational research.

The supercomputer consists of:

• 3 Appro 1U Quad XtremeServer Linux clusters
  
• 2,016 quad socket nodes and 16,128 cores based on the latest AMD dual-core processors
  
• Voltaire Grid Director ISR 9288 multi-service switches with DDR InfiniBand
  
• Grid Switch ISR 9024 edge switches with DDR InfiniBand

"With the Peloton project, the Laboratory will deploy InfiniBand interconnect into large scale production resources for the first time. Voltaire's DDR solution will significantly increase the bandwidth and messaging rate required to meet the ever-increasing demand for the large scale simulations indispensable to advancing all scientific disciplines."

Mark Seager, ICCD ADH for Advanced Technology, LLNL



Trinity College Deploys Voltaire Grid Backbone Solutions
Dublin, Ireland-based Trinity College has deployed an IBM xSeries clustered supercomputer with Voltaire Grid Backbone Solutions to advance its computational science and research initiatives. The Trinity supercomputer is used to conduct research in a variety of areas that include the physical sciences, biomolecular sciences and computer science. The 356-node cluster is the largest IBM-powered InfiniBand cluster deployed to date.

The supercomputer which ranked 226th on the November 2005 Top500 list, consists of 356 IBM eServer model 326 servers with dual AMD model 250 Opteron processors and Voltaire Grid Director ISR 9288 multi-service switches for the 10 Gbps InfiniBand connectivity. The system also uses IBM General Parallel File System (GPFS), a high-performance shared disk file system that provides fast data access from all of the nodes in the cluster

"This new supercomputer will contribute to important progress in computational science in Ireland. Interconnecting the IBM platforms with Voltaire's 288 port InfiniBand switches gives us a high-performance, scalable supercomputer that allows us to tackle larger and more complex simulations than have been possible before."

Dr. Graeme Watson
Director of the Trinity Centre for High Performance Computing

Click here to read more about Trinity's supercomputer



Voltaire Improves Performance & Server Utilization for Sandia's "Catalyst" Cluster
A long-time customer of Voltaire, Sandia National Laboratories selected the Voltaire Grid Director ISR 9288, a 288 port 4X (10 Gbps) multi-service switch, to provide the high performance interconnect for supercomputer-class clusters.

Sandia is using the Grid Director ISR 9288, to improve the performance and server utilization of its 128 node P4 Xeon "Catalyst" cluster, which ranked no. 334 in the Top500 list of the world's fastest supercomputers published in November 2004.

"After considering products from several InfiniBand vendors, we selected Voltaire's solutions on the strength of their performance, maturity and scalability. For the past year, Sandia has used our InfiniBand high performance cluster for R&D and science application runs. We've been impressed with the results. A switch with 288 InfiniBand ports in a single chassis greatly simplifies the process of building, scaling and managing large scale clusters of hundreds to thousands of nodes."

Matt Leininger
Computational Scientist
Sandia National Laboratories



CERN Picks Voltaire for Worldwide Grid
CERN and its academic partners around the globe are pioneering a worldwide Grid called the LHC Computing Grid (LCG) which is expected to be the largest data-intensive application of the decade. This computing Grid will be used to analyze data from the Large Hadron Collider -- the biggest scientific instrument on the planet -- sifting through petabytes of particle collision data, looking for clues to the origins of the Universe. The LCG will rely on clusters of computers at CERN and at hundreds of partner universities and institutes. CERN selected Voltaire InfiniBand interconnect solutions for the project because they provide four essential features: low latency, very high bandwidth, low CPU overhead and powerful connectivity to storage.

• Voltaire Grid Director ISR 9600 Switch
  
• Voltaire Grid Switch ISR 6000
  
• Voltaire Fibre Channel Routers

"The high bandwidth, low CPU overhead and scalability of Voltaire's InfiniBand solutions are particularly interesting to the computer clusters used for high performance analysis in the LCG. With InfiniBand, data can be streamed into our CERN opencluster very quickly with minimal loss of CPU cycles, so we can retain as many cycles as possible for the data analysis itself."

Sverre Jarp
Chief Technology Officer,
CERN openlab



Shanghai University Hits Top500 with Voltaire 288 Port Switch
Shanghai University selected the Voltaire Grid Director ISR 9288, a 288 port 4X (10 Gbps) multi-service switch to deliver the ultimate interconnect performance for its supercomputer cluster. Using the single high port count InfiniBand switch, the cluster achieved extremely high bandwidth and low latency out of the box and was ranked no. 209 in the November 2004 Top500 list.

• Voltaire Grid Director ISR 9288, a 288 port 4X (10 Gbps) multi-service switch
  
• 174 HP ProLiant DL360 servers

Voltaire Delivers Performance for U. Minnesota's 64-bit Linux Cluster
To achieve supercomputing-class performance, the University of Minnesota Supercomputing Institute deployed Voltaire Grid Backbone solutions as part of its mid-range technical computing system of 64-bit Linux OS-based clusters.

Running a full slate of supercomputing applications, the high performance shared memory computing solution consists of:

• Voltaire Grid Switch ISR 9024
  
• 8, 16-processor SGI Altix 350 mid range systems

CEA Picks Voltaire for High Performance & Low Latency
CEA, the Commissarriat a l'Energie Atomique in France selected Voltaire Grid Backbone solutions as the high bandwith, low latency interconnect to power a 60-node cluster. CEA uses the Voltaire Grid Director ISR 9600 multi-service switch and HP ProLiant DL585 4-way Opteron servers.

• Voltaire Grid Director ISR 9600 Switches
  
• 60 HP ProLiant DL585 4-way Opteron servers

Alabama Supercomputes with Voltaire
The Alabama Supercomputer Center selected Voltaire Grid Backbone solutions for its five-node, 56-processor cluster of SGI Altix 350 systems. Sets of 16 CPUs are grouped together into shared memory nodes and the Voltaire Grid Switch ISR 9024 is used for message passing between the nodes.

• Voltaire Grid Switch ISR 9024
  
• 5 SGI Altix 350 mid range systems

Voltaire Drives Maverick to Top500
The ERC at Mississippi State University is leveraging Voltaire's advanced Grid Backbone solutions to support "Maverick," a 192-node, diskless IBM server cluster - the largest diskless InfiniBand server cluster built to date.

• Voltaire Grid Director ISR 9600 Switch
  
• GridVision™ InfiniBand Fabric Management Software
  
• 192 IBM eServer xSeries x335

"With the power, performance and scalability of Voltaire's InfiniBand switches and the robust functionality of the VoltaireVision management software, we are well-enabled to achieve significant throughput and performance gains."

Trey Breckenridge
HPC Resources and Operations Administrator
Mississippi State University ERC



Voltaire Improves Performance of Red Storm Supercomputer
Sandia National Laboratories uses Voltaire Grid Backbone solutions for a large data and visualization cluster that supports the Red Storm supercomputer. This cluster was named to the November 2004 Top500 list.

• Voltaire Grid Backbone ISR 9600 Switch
  
• Voltaire IP routers
  
• Voltaire Fibre Channel routers
  
• Voltaire HCA400 host channel adapters with Voltaire GridStack
  
• 136 Dell PowerEdge Servers

"After considering products from several InfiniBand vendors, we selected Voltaire's solutions on the strength of their performance, maturity and scalability."

Milton Clauser, Ph. D.
Sandia National Laboratories



Voltaire Interconnect Solutions Power Ohio's HPC Linux Cluster
OSC selected Voltaire Grid Backbone solutions as the high bandwidth, low latency interconnect for a 128 node cluster of industry-standard servers. OSC uses two Voltaire Grid Director ISR 9600 switches, embedded with GridVision™ InfiniBand Fabric Management Software, to power the 128 node cluster based on Intel Pentium 4 machines. OSC's cluster was named to the November 2004 Top500 list.

The cluster supports research in the areas of bioinformatics, biomedical applications, computational fluid dynamics, computational reactive mechanics and OSC's own advanced MPI-based applications research.

"We've used our 128 node Voltaire InfiniBand cluster to port PVFS2 to run over InfiniBand. PVFS2 is used in production to provide a high-performance parallel I/O capability to our users for their data-intensive applications. Running PVFS2 over InfiniBand further increases data throughput and lowers metadata access latencies."

Dr. Pete Wyckoff
Research Scientist
The Ohio Supercomputer Center