November 20, 2008
Bicycle-powered computer simulations highlight supercomputing industry progress, demonstrate SiCortex energy efficiency
AUSTIN, Texas, Nov. 20 -- One hundred years ago, Lewis Fry Richardson completed the first true scientific simulation, a two-dimensional model to predict the behavior of a dam under stress. This project involved a team of human "computers" performing slide-rule calculations for a mind-and-finger-numbing period of three years. This week, a team of humans performed a similar simulation, this time by pedaling bicycles to power a supercomputer. The demonstration was held at SC08 in Austin, Texas, by students from Purdue University and seven cyclists from Mellow Johnny's Bike Shop, a venture partially owned by Lance Armstrong. The Purdue students used a SiCortex SC1458 high-productivity computer to complete a three-dimensional dam-break simulation in less than 20 minutes. "Supercomputing has come a long way over the last hundred years," said SiCortex president and CEO Chris Stone.
Another metric of this progress is the computer that performed the simulation. The SiCortex computer is the world's most energy efficient, producing answers to complex scientific computing challenges as quickly as conventional cluster computers, using a fraction of a cluster's energy requirement -- small enough, in fact, to be powered by bicyclists. The SiCortex computer used in Purdue's demonstration comprised 648 processors designed specifically for scientific computing, linked together by a warp-speed communication fabric. This optimized design, free of the superfluous chips used by conventional, hot-running clusters, required no special cooling as it 'broke the dam'. "Indeed, only the cyclists appeared to generate heat," stated one observer.
"Richardson not only pioneered the science of mathematical simulation, he also demonstrated the importance of parallel processing as a means to achieve answers," said Stone. "The talented Purdue team has appropriately honored Richardson's achievement by using the world's most energy-efficient parallel processing system -- reducing 'cycle time' to a matter of minutes."
With the help of the SiCortex SC1458, Purdue also showcased the "super power" of supercomputers during the highly competitive SC08 Cluster Challenge. The competition, a personification of the great progress made in scientific computing, challenged educational institutions to build a system and then run simulated data in a range of scientific applications, including genetics and designing jet airplanes. Outside of the official competition, the Purdue students hit another supercomputing milestone by being the first team to ever achieve a teraflop of computing after they completed the Cluster Challenge requirements. The results of the Cluster Challenge are expected to be released later today.
Headquartered near Boston, Mass., SiCortex, Inc. makes the world's most energy-efficient high-productivity computers. Its proven architecture was designed from the silicon up to provide breakthrough delivered performance at the lowest power consumption in the industry. SiCortex computers scale from 72 to 5,832 processors running Linux and other open-source codes, in packages ranging from deskside to departmental to data center. SiCortex systems are the compute-power behind some of the most important research initiatives at the country's government agencies, national laboratories and academic institutions. For more information, visit http://www.sicortex.com.
Source: SiCortex, Inc.
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?