November 15, 2004
Chelsio Communications Inc exhibited its 10Gb Ethernet technology critical to the newly established world record for Ethernet LAN speed and distance. Chelsio's 10GbE leadership was showcased with a wide range of demonstrations and exhibits highlighting the high performance and low latency T110 10GbE Protocol Engine at the 2004 Supercomputing Conference.
Chelsio played a critical and central role in the record breaking 10Gigabit Ethernet link for the transmission of data from the University of Tokyo to the CERN research center in Geneva, Switzerland. Chelsio's T110 delivered sustained 7.57 Gbps throughput running standard 1500-byte Ethernet packets over a single TCP connection across the 18,500 km link using a uni-processor AMD Opteron system on each end of the connection. In the single stream stress test, 7.5 Gbps was sustained for nearly 7 hours, a critical capability for long-distance data backup applications.
"Achieving the world record in our CERN-Tokyo experiments required unique capabilities available with Chelsio's T110 protocol engine" said Kei Hiraki of the University of Tokyo. "The programmability and flexibility of the T110 was critical to our success, and this level of intelligence must be a standard feature of 10G adapters for data transfer on long fat-pipe networks."
"The goal of 10Gb Ethernet is simple: drive the convergence of networking, storage and cluster computing interconnect to the ubiquitous Ethernet, eliminating the need for proprietary and niche networking technologies. Chelsio is driving this convergence by matching the performance attributes, such as high throughput and low latency, of competing interconnects such as Fibre Channel and Infiniband," said Asgeir Eiriksson, chief technology officer for Chelsio. "To achieve this, the TCP termination function must be offloaded from the host CPU to a programmable TCP Offload Engine such as Chelsio's Terminator processor. Industry leading throughput, low latency and multiple other benefits of our Terminator architecture are available only from Chelsio's T110 adapter."
"Chelsio's T110 protocol engine consistently delivered very high performance during our testing, reliably sustaining a single stream TCP transfer rate of 7.5 gigabits per second for most of our 7-hour stress test," said Catalin Meriosu of CERN. "I was impressed by how quickly Chelsio's TCP offload engine regained top transfer rate in the presence of isolated packet loss, which demonstrates the stability and robustness of their product."
"Our success with the Tokyo to Geneva speed record is a testament to the flexibility and robustness of Chelsio's protocol offload adapter," said Kianoosh Naghshineh, president and CEO of Chelsio. "Chelsio's T110 protocol engine has proven its high performance and resiliency through several enterprise OEM qualifications, and has now demonstrated these capabilities in challenging topologies such as this transcontinental link."
10/30/2013 | Cray, DDN, Mellanox, NetApp, ScaleMP, Supermicro, Xyratex | Creating data is easy… the challenge is getting it to the right place to make use of it. This paper discusses fresh solutions that can directly increase I/O efficiency, and the applications of these solutions to current, and new technology infrastructures.
10/01/2013 | IBM | A new trend is developing in the HPC space that is also affecting enterprise computing productivity with the arrival of “ultra-dense” hyper-scale servers.
Ken Claffey, SVP and General Manager at Xyratex, presents ClusterStor at the Vendor Showdown at ISC13 in Leipzig, Germany.
Join HPCwire Editor Nicole Hemsoth and Dr. David Bader from Georgia Tech as they take center stage on opening night at Atlanta's first Big Data Kick Off Week, filmed in front of a live audience. Nicole and David look at the evolution of HPC, today's big data challenges, discuss real world solutions, and reveal their predictions. Exactly what does the future holds for HPC?