U.S. regains supercomputer lead
- 16 July, 2012 10:10
- Comments
The U.S. is once again home to the world's most powerful supercomputer, rebounding after it was knocked out of the top spot by China two years ago and Japan last year.
The latest Top500 list of supercomputers, released last month, also marks a return of European technology in force, as the continent accounted for four of the first 10 systems on the list, and reveals that these ever-powerful computers are increasingly used for commercial purposes.
Even though U.S. technology has dominated the Top500 rankings since researchers first compiled the list in 1993, the two-year absence from the top spot had hit a national nerve.
President Barack Obama mentioned China's top-ranked supercomputer in two separate speeches, including last year's State of the Union address, and Secretary of Energy Steven Chu, a Nobel Prize winner in physics, warned that America's innovation leadership was at risk.
Though competitive threats to U.S. technology leadership remain -- for example, China launched its fourth manned space mission in June, just months after the U.S. ended its shuttle program -- the U.S. is back on top of the supercomputer world, at least for now.
The Linux-based IBM Sequoia system that tops the latest list is powered by Power BQC 16-core processors running at 1.6GHz, and mostly relies on architecture and parallelism, not Moore's Law, to achieve speeds of 16.32 sustained petaflops. Sequoia, an IBM BlueGene/Q, is installed at the U.S. Department of Energy's Lawrence Livermore National Laboratory.
"We're at the point where the processors themselves aren't really getting any faster," said Michael Papka, deputy associate director for computing, environment and life sciences at the DOE's Argonne National Laboratory, which runs a similar IBM system that placed third on the list. "Moore's Law is generally slowing down and we're [getting faster speeds] by parallelism."
"The classic trick of waiting for Moore's Law to come along and help you out really doesn't exist anymore," added David Turek, vice president of exascale computing at IBM.
Both the Sequoia supercomputer and the Mira system housed at the Argonne lab are among the first IBM supercomputers to use water-cooling technology.
The most striking thing about the list, though, to Jack Dongarra, a professor of computer science at the University of Tennessee and a leader of the Top500 program, is that more than half of the machines on it aren't deployed in research, in academic settings or by government.
"More than half are used by industry," he said. "Industry gets it. These machines are important; they can provide some competitive advantage."
Europeans, in particular, are moving aggressively to build out commercial supercomputing capability, despite all the troubles their economies are facing. "The Europeans weren't keeping pace," Dongarra noted. "Today, we see a resurgence."
The European machines in the top 10 -- two in Germany, one in Italy and one in France -- are new, he said.
IBM made 213 of the systems on the list, including five of the top 10. Hewlett-Packard was the second most well-represented computer maker, with 141 of the 500 systems. Nearly 75% of the systems listed run on Intel processors, and 13% use AMD chips.
Mikael Rickns of the IDG News Service contributed to this story.
This version of this story was originally published in Computerworld's print edition. It was adapted from an article that appeared earlier on Computerworld.com.
Read more about hardware in Computerworld's Hardware Topic Center.
Recommended
Join the CIO Australia group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.
- Bookmark this page
- Share this article
- Got more on this story? Email CIO
- Follow CIO on twitter
-
Why change management doesn’t work
-
Larry Page wants to see your medical records
-
Dual-Persona Smartphones Not a BYOD Panacea
-
After two-year hiatus, EFF accepts bitcoin donations again
-
CIOs struggle to deliver timely mobile business apps: survey
-
HP Helps NEC Reduce Network Management Costs and Gain Efficiencies
NEC wanted to reduce network management costs, while increasing network visibility, decreasing mean-time-to-repair, improving stability and mitigating the risk of downtime. Download today to hear from Cameron Craig, Senior department manager of NEC on what approach they took and why. -
Endpoint Security and Virtualisation
Besides form factor, virtual systems are not really that different than physical systems. They both use the same operating systems and applications. They both present users with computing resources such as RAM and hard drives. Consequently, the ability to exploit vulnerabilities in a physical environment will present a significant threat to virtualised environments as well. This paper examines the different endpoint security methods for virtualised environments and presents how Endpoint Protection security provides optimal performance, protection and manageability. -
Agentless Security for Virtual Environments
Virtualised datacentres, desktops, and cloud computing should be secured by the same strong protection technologies as physical machines. However, traditional agent-based solutions that are not architected for virtualisation can result in a number of significant operational security issues. Find out more about the first agentless security platform solution.
Get exclusive access to Invitation only events CIO, reports & analysis. 
