The Leading Source for Global News and Information Covering the Ecosystem of High Productivity Computing
June 18, 2008
Mellanox InfiniBand Accelerates Top Supercomputer
20Gb/s ConnectX provides the foundation for interconnecting five of the top ten computers
DRESDEN, Germany, June 18 -- Mellanox Technologies, Ltd., a leading supplier of semiconductor-based, server and storage interconnect products, today announced that its ConnectX 20Gb/s InfiniBand adapters and switch systems based on its InfiniScale switch silicon provide the leading, scalable, low-latency, high-performance interconnect for the #1 fastest supercomputer on the Top500, National Nuclear Security Administration’s (NNSA) Roadrunner, the world’s first petaflop supercomputer. Built by IBM and housed at NNSA’s Los Alamos National Laboratory in New Mexico, Roadrunner’s performance is more than double the performance of the next leading contender on the latest TOP500 list of supercomputers. The computer will be used to analyze national security threats, test nuclear stockpiles, run annual testing of various nuclear weapons systems, and investigate long-term climate change, astronomy, and human genome science. By providing the underlying fabric technology for the world’s most powerful supercomputer, Mellanox continues to demonstrate its technical proficiency, proven reliability and scalability, and performance leadership.
“Mellanox’s interconnect products provide leading technology and device architecture designed for scalable high-performance and low-power consumption, and are proven to optimize the world's most demanding computing and storage applications,” said Thad Omura, vice president of product marketing at Mellanox Technologies. “We are pleased to have our technology at the foundation of five of the Top10 systems in the TOP500 list, and in nearly half of the Top100 systems. Our newly announced 40Gb/s InfiniBand solutions will continue to support the ever-increasing performance needs of world leading research, educational centers and commercial datacenters.”
Published twice a year and publicly available at Top500.org, the TOP500 list ranks the most powerful computer systems according to the Linpack benchmark rating system and is a respected industry report which indicates usage trends in computing and interconnect solutions.
Highlights of InfiniBand usage on the June 2008 TOP500 list include:
- Mellanox InfiniBand interconnect products occupy 5 of the top 10 most prestigious positions.
- All InfiniBand-based clusters utilize Mellanox switch silicon and over 95% integrate adapter products based on Mellanox silicon.
- InfiniBand is the only industry standard interconnect used in the Top58 (29 total supercomputers) -- all based on Mellanox adapter cards and switch silicon.
- Mellanox InfiniBand interconnect products present in the TOP500 are used by a diverse list of applications, from large-scale high-performance computing to commercial technical computing and enterprise datacenters.
- The average efficiency of all reported InfiniBand-based supercomputers is 73% (significantly higher than 53% for Gigabit Ethernet supercomputers) reducing power, cooling, cost, floor space and total cost of ownership.
- InfiniBand is the most used interconnect in the Top100 supercomputers with 49 clusters, more than 6X the number of Gigabit Ethernet based clusters (8), and nearly 5X the number of proprietary high speed cluster interconnects.
- The total number of InfiniBand connected CPU cores on the list has grown from 340,000 in November 2007 to 607,000 in 2008 (78% half-yearly growth). This growth is nearly 2X the growth in the total TOP500 list, which highlights InfiniBand’s world leading scalability.
- InfiniBand is the only growing clustered interconnect in the Top100 with nearly a 30% growth rate compared to Nov 2007 and 180% compared to June 2007.
- The total performance of InfiniBand based systems has grown from 2PFlops to 6PFlops (196% yearly growth). This growth is 3X the growth in the total TOP500 list, which highlights the increasing demand for InfiniBand to maximize computing resources and performance.
- The number of InfiniBand-based supercomputers is more than 7 times the number of supercomputers using other high-speed interconnects combined (> 1 Gb/s).
- The entry-level for the TOP500 list is 8.9TFlops, 150% higher than the 5.9TFlops necessary to be on the November 2007 list, indicating the explosive market demands for more efficient and scalable computing enabled by Mellanox’s interconnect products.
About Mellanox
Mellanox Technologies (NASDAQ: MLNX; TASE: MLNX) is a leading supplier of semiconductor-based, interconnect products to world-class server, storage, and infrastructure OEMs servicing Fortune 500 datacenters, the world’s most powerful supercomputers, and mission critical embedded applications. The company’s Virtual Protocol Interconnect (VPI) enables standard communication protocols to operate over any converged network (InfiniBand, Ethernet, Data Center Ethernet) with the same software solution. Utilizing proven networking, clustering, storage, virtualization and RDMA acceleration engines, VPI optimizes application performance, power consumption, workload agility, and total system efficiency while future-proofing IT infrastructure.
-----
Source: Mellanox Technologies
Sponsored Links
Interview: Appro CEO Shares HPC Vision
Appro CEO Daniel Kim provides a glimpse into Appro's vision and opportunities for its supercomputer and high-performance cluster solutions.
RSS Feeds
Feeds by Topic
- Applications
- Developer Tools
- Interconnects
- Middleware
- Networks
- Processors
- Storage
- Systems
- Visualization
Feeds by Industry
Feeds by Content Type
Feature Articles
Computed Tomography Software Taps Into NVIDIA GPUs
Minnesota-based North Star Imaging, a firm that specializes in industrial X-rays for nondestructive testing and analysis, is employing NVIDIA GPUs to accelerate 3D renderings in their CT (computed tomography) software. Julien Noel, the company's CT product manager, says the exceptional computational power afforded by CUDA and Tesla hardware is increasing customer productivity and transforming their workflow.
Read More...
The Next Big Thing in Humanities, Arts and Social Science Computing: 18thConnect
For the humanities scholar who may have only recently mastered library and archival finding aids beyond the archaic card catalog, the possibility of retrieving source materials at the flash of a keystroke (well maybe a few...) is very heady stuff.
Read More...
HPC Clouds -- Alto Cirrus or Cumulonimbus
The "cloud" model of exporting user workload and services to remote, distributed and virtual environments is emerging as a powerful computing paradigm. Yet, one domain that challenges this model in its characteristics and needs is high performance computing.
Read More...
Top Headlines
Dawning 6000 to Use Chinese-Made Loongson Processor
Nov 28 | People's Daily Online | Currently under development, the Dawning 6000 HPC system will be based on the Chinese-made "Loongson" microprocessor. Read more...
Can Supercomputers Help Save the Economy?
Nov 27 | Computerworld | The use of supercomputers to increase the industrial might of the U.S. has amounted to little more than an asterisk from a financial standpoint in both the federal budget and the economy as a whole. Read more...
IBM to Establish 'Collaboratory' in Dublin
Nov 26 | Science Business | IBM is getting ready to set up a supercomputing research “collaboratory” in Dublin, Ireland. Read more...
Texan Prof Sees Big Future for Graphene Storage
Nov 25 | The Register | A Rice University professor believes that his proposed graphene arrays could be many times denser and faster than existing storage tech, and they'd be more reliable too. Read more...
Super Micro Computer: A One-Man, or at Least One-Family, Powerhouse
Nov 24 | The New York Times | Server maker Super Micro Computer lives by two principles: give customers what they want, and do it as fast as humanly possible. Read more...
Multimedia
Video White Paper: Architecting a Better Network Storage Solution
BlueArc's Titan architecture represents an evolutionary step in file servers by creating a hardware-based file system that can scale bandwidth, IOPS, and overall data capacity well beyond conventional software-based devices. With its ability to virtualize a massive storage pool of up to four usable petabytes of tiered storage, Titan can scale with growing data requirements, offering a competitive advantage for businesses, researchers, or other enterprises seeking to better manage data growth while still ensuring optimal performance.
Newsletters
Stay informed! Subscribe to HPCwire email Newsletters.
Get updates and insights on the High Productivity Computing industry delivered driectly to your inbox.
HPC Job Bank
-
Sr. Systems Developer/Engineer
Ohio State University -
HPC System Administration Specialist
Argonne National Laboratory - Visit the HPCwire Job Bank
Featured Events
-
February 24-27, 2009
TradeTech Architecture 2009
London England
-
April 13-16, 2009
Symposium 2009: LEAN, CLEAN, and GREEN
New York , NY
USA