SpyderByte.com ;Technical Portals 
      
 News & Information Related to Linux High Performance Computing, Linux Clustering and Cloud Computing
Home About News Archives Contribute News, Articles, Press Releases Mobile Edition Contact Advertising/Sponsorship Search Privacy
HPC Vendors
Cluster Quoter (HPC Cluster RFQ)
Hardware Vendors
Software Vendors
HPC Consultants
Training Vendors
HPC Resources
Featured Articles
Cluster Builder
Beginners
Whitepapers
Documentation
Software
Lists/Newsgroups
Books
User Groups & Organizations
HP Server Diagrams
HPC News
Latest News
Newsletter
News Archives
Search Archives
HPC Links
ClusterMonkey.net
Scalability.org
HPCCommunity.org

Beowulf.org
HPC Tech Forum (was BW-BUG)
Gelato.org
The Aggregate
Top500.org
Cluster Computing Info Centre
Coyote Gultch
Dr. Robert Brown's Beowulf Page
FreshMeat.net: HPC Software
SuperComputingOnline
HPC User Forum
GridsWatch
HPC Newsletters
Stay current on Linux HPC news, events and information.
LinuxHPC.org Newsletter

Other Mailing Lists:
Linux High Availability
Beowulf Mailing List
Gelato.org (Linux Itanium)

LinuxHPC.org
Home
About
Contact
Mobile Edition
Sponsorship

Latest News

Mellanox Powers World-Class Sockets Performance Over 20Gb/s InfiniBand Links
Posted by Kenneth Farmer, Monday September 26 2005 @ 07:50AM EDT

SANTA CLARA, CA -- 09/23/2005 -- Mellanox Technologies Ltd, the leader in business and technical computing interconnects, announced Zero-Copy (ZCopy) Sockets Direct Protocol (SDP) running over InfiniBand fabrics can generate double the data throughput between clustered server nodes, while reducing the overall CPU utilization by up to a factor of ten when compared to other solutions. Mellanox InfiniBand fabric products are the only low-latency, high-performance 20Gb/s interconnect solutions that offload transport processing in hardware. The wide range of mainstream sockets-based applications that run over a InfiniBand connected server clusters using ZCopy SDP obtain industry-leading performance due to optimal CPU utilization, maximum interconnect throughput, and minimal latency. Benchmarks show that a 20Gb/s server-to-server InfiniBand link supports over 1360MB/s of data throughput utilizing ZCopy SDP -- more than double the performance of older sockets-based implementations limited by CPU processing bottlenecks.

"With 20Gb/s InfiniBand server nodes in the market today and 40Gb/s capabilities around the corner, an efficient communication protocol that scales with fabric speeds, rather than CPU horsepower, is vital to improve clustered computing performance and efficiency," said Dror Goldenberg, Senior Member of Mellanox Technologies' Architecture Team. "ZCopy SDP enables the plethora of mainstream sockets-based applications to efficiently utilize zero buffer copy and transport offload capabilities of the InfiniBand fabric for generations to come."

ZCopy SDP implementation and benchmark results will be discussed at the Remote Direct Memory Access (RDMA): Applications, Implementations, and Technologies (RAIT 2005) Workshop to be held on September 26th, 2005, at the Burlington Marriott near Boston, Massachusetts. Dror Goldenberg from Mellanox will present a paper entitled "Transparently Achieving Superior Socket Performance Using Zero Copy SDP over 20Gb/s InfiniBand Links." The research paper highlights how pipelining communication over multiple connections via ZCopy SDP ultimately drives the highest performance and efficiency of server-to-server sockets traffic using 20Gb/s InfiniBand links. ZCopy SDP open-source drivers will be widely available in 4Q2005.

Visit the Mellanox Booth at the Cluster 2005 Conference

< Scalable Systems integrates Sun N1GE into the Rocks Cluster Distribution | ClearSpeed Claims Fastest Chip Crown >

 

Affiliates

Cluster Monkey

HPC Community


Supercomputing 2010

- Supercomputing 2010 website...

- 2010 Beowulf Bash

- SC10 hits YouTube!

- Louisiana Governor Jindal Proclaims the week of November 14th "Supercomputing Week" in honor of SC10!








Appro: High Performance Computing Resources
IDC: Appro Xtreme-X Supercomputer Blade Solution
Analysis of the Xtreme-X architecture and management system while assessing challenges and opportunities in the technical computing market for blade servers.

Video - The Road to PetaFlop Computing
Explore the Scalable Unit concept where multiple clusters of various sizes can be rapidly built and deployed into production. This new architectural approach yields many subtle benefits to dramatically lower total cost of ownership.
White Paper - Optimized HPC Performance
Multi-core processors provide a unique set of challenges and opportunities for the HPC market. Discover MPI strategies for the Next-Generation Quad-Core Processors.

Appro and the Three National Laboratories
[Appro delivers a new breed of highly scalable, dynamic, reliable and effective Linux clusters to create the next generation of supercomputers for the National Laboratories.

AMD Opteron-based products | Intel Xeon-based products



Home About News Archives Contribute News, Articles, Press Releases Mobile Edition Contact Advertising/Sponsorship Search Privacy
     Copyright © 2001-2013 LinuxHPC.org
Linux is a trademark of Linus Torvalds
All other trademarks are those of their owners.
    
  SpyderByte.com ;Technical Portals