Visit blinkin's column >>

BLINKIN

News Junkie
Add To Watchlist
Articles Posted: 23; Links Seeded: 1347
Member Since: 5/2008Last Seen: 10/03/2009

USA Supercomputer Splurge

Credit: LeRoy N. Sanchez, Records Management, Media Services and Operations (Los Alamos National Lab)

Credit: LeRoy N. Sanchez, Records Management, Media Services and Operations (Los Alamos National Lab)

Purdue computer technicians put the finishing touches on Steele, a new supercomputer that is among the largest in the world. Staff members at the university were challenged to build the supercomputer in a day, but finished the job by lunchtime. (Purdue News Service photo/David Umberger)

advertisement

"Its exciting to think that we can get the best brains from super-computing and from mass computing and bring those together and make great progress in the decade ahead."

- Bill Gates, at the Super-Computing 2005 Conference in Seattle.

3 years after Bill Gates articulated his vision, progress towards the goal has been both exciting and phenomenal. A new generation of compact,
ultra powerful, energy efficient supercomputers are being installed across the country.

Back in March, Sun Systems and The Texas Advanced Computing Center debuted the most powerful supercomputer ever built, with over
500 teraflops of computing power.

Last week, An American military IBM supercomputer at Lawrence Livermore National Labratory in California surpassed a computing milestone by processing more than 1.026 quadrillion calculations per second, or more than a petaflop.

This week, U.S. Secretary of Energy Samuel Bodman announced that the new 'Roadrunner' IBM supercomputer at Los Alamos lab in New Mexico is the first to achieve a petaflop of sustained performance, making it the world's fastest supercomputer. This hybrid computer runs Linux by Red Hat, and is being used to mimic complex human nuerological processes involving the human visual cortex.

"This new supercomputer demonstrates a commitment to achieve a major advance in technological capability that will help enable scientists and businesses solve the most challenging problems," said Bill Zeitler, senior vice president, IBM Systems and Technology Group.

Purdue University also installed a new energy efficient supercomputer in May that uses 40% less energy than standard systems. Gerry McCartney, vice president for information technology and chief information officer at Purdue, explains; "Everyone understands in this industry that energy is an issue.
We are trying to find solutions with a new company and a new product because that's where Purdue belongs: with people who are looking for solutions."

A new generation of extremely fast supercomputers raises new possibilities. To plan for this future,
a special panel discussion on "RoadRunner — the First Petaflop/s System in the World and Its Impact
on Supercomputing" will be held June 18th in Dresden, Germany, at the International Supercomputing Conference.

Expect more Universities, labaratories and businesses across the country to install these new supercomputers. The age of the supercomputer has begun.

Copyright 2008 by blinkin. This material may not be reproduced, rebroadcast, or distributed. Photos are copyright of their authors, see pic for credit.

SOURCES:
Microsoft aims for super computing: Gates
Sun Unveils Most Powerful Supercomputer
Military Supercomputer Sets Record
IBM to Build World's First Cell Broadband Engine Based Supercomputer
Los Alamos National Laboratory
U.S. Department of Energy's New Supercomputer is Fastest in the World
Scientists using Roadrunner supercomputer to mimic complex brain mechanisms
Purdue's supercomputer uses less energy
Purdue supercomputer unboxed and built by lunchtime
International Supercomputing Conference to Host First Panel Discussion on Breaking the Petaflop/s Barrier

Additional Reading:

At Los Alamos, Two Visions of Supercomputing
You can even build your own supercomputer from 8 Playstation 3's .
From India, 2003, comes a theory that The Universe is a Massive Supercomputer
Our Energy Future

More Articles and Seeds From blinkin

  • 3 Votes
  • Enjoy this article? Help vote it up the 'Vine.

Back To Top

Published to:

What's this?
Who's leading the conversation?
This visualization below allows you to see the impact that each user has on the current conversation. The top row contains the group of users who have had the most impact, the 2nd row the group of users who have had the 2nd most impact (et cetera). Users with similar impact are grouped together, and the average score of the group is shown to the left of the group. The author of the article is also shown on the left, in their corresponding group. Each user's score is based on the number of comments the user has made plus the number of votes their comments have received. The scores are calculated relative one another, so while their absolute value is not particularly important, their relative difference does indicate a larger difference in impact on the conversation.
1.1
{"commentId":1991451,"authorDomain":"inthezone1"}

From Wikipedia, the free encyclopedia (Redirected from Teraflop)
Jump to: navigation, search
Computer Performance
Name flops
megaflop 106
gigaflop 109
teraflop 1012
petaflop 1015
exaflop 1018
zettaflop 1021
yottaflop 1024

In computing, FLOPS (or flops or flop/s) is an acronym meaning FLoating point Operations Per Second. The FLOPS is a measure of a computer's performance, especially in fields of scientific calculations that make heavy use of floating point calculations, similar to instructions per second.

{"commentId":1991451,"threadId":"292223","contentId":"1572498","authorDomain":"inthezone1"}
  • 1 vote
Reply#1 - Wed Jun 18, 2008 12:23 AM EDT
{"commentId":1995537,"authorDomain":"jg2090"}

That's right, inthezone. They measure the performance of supercomputers using a program called LINPACK or LAPACK, a program for performing numerical linear algebra. The highest score measured using the LINPACK benchmark suite is the number which is used to rank the computers, measured in trillions of floating point operations per second, i.e. Teraflops.

The Road Runner Supercomputer at Los Alamos received an RMAX of 1026 Teraflops, or 1.026 Petaflops, using 122400 processor cores.

The next closest supercomputer had an RMAX of 478.2 Teraflops, using 212992 processor cores. This is the IBM BlueGene/L supercomputer at Lawrence Livermore National Laboratory in California.

SOURCE:

Computer cluster

{"commentId":1995537,"threadId":"292223","contentId":"1572498","authorDomain":"jg2090"}
    Reply#2 - Wed Jun 18, 2008 1:48 PM EDT
    {"commentId":1996021,"authorDomain":"ryan-jas"}

    When they said that the units use 40 percent less energy, does that mean entire system uses less energy? From I've heard supercomputers require massive cooling systems that in some cases require more energy than the computer units themselves. Granted that these computers are running super duper fast I think we can safely assume that cooling them is much more of a concern, so does this 40 percent reduction just measure the computers consumption or does it also factor in the cooling side of the equation?

    {"commentId":1996021,"threadId":"292223","contentId":"1572498","authorDomain":"ryan-jas"}
      Reply#3 - Wed Jun 18, 2008 2:29 PM EDT
      {"commentId":1997576,"authorDomain":"jg2090"}

      First, from my last post above, you can see the difference between the #1 and the #2 supercomputers. Roadrunner achieves twice the Teraflops using a system with 40% fewer physical processor cores. That alone would net an energy savings.

      IBM is building a new data center for Rackforce in Canada. Rackforce indicates that their old data center had a PUV or Power Usege Effectivness rating of 3.0. Their new data center has a PUV of 1.3. Watch their video.

      IBM is also using heat exchangers to help with cooling costs. One such system they built in Denver, Co, takes advantage of the climate in Denver:

      When exterior temperature and humidity levels are favorable, the new data center's technology switches to free-cooling mode -- utilizing a water economizer to dramatically reduce energy consumption.

      Variable-speed pumps and motors were also installed in the air conditioning systems to balance the cooling capacities to the actual load. This further reduces energy usage and costs, and water taps were installed throughout the facility, providing flexibility for water-cooled IT equipment -- like the IBM Power 6 technology -- as well as today's air-cooled profiles. The center also uses low-sulfur fuels to reduce emissions from backup generators.

      IBM Project Big Green Tackles Global Energy Crisis

      IBM scientists have developed a method to cool computer chips that have circuits and components stacked on top of each other with tiny rivers of water. IBM demonstrated a prototype that integrates the cooling system into the three dimensional (3-D) chips by piping water directly between each layer in the stack. These so-called 3-D chip stacks -- which take chips and memory devices that traditionally sit side-by-side on a silicon wafer and stacks them together on top of one another -- presents one of the most promising approaches to enhancing chip performance beyond its predicted limits, while simultaneously reducing the energy consumed by data centers.

      IBM is innovating every step in the data center process to maximize efficiency and to reduce cost. This is a company that is solving problems before they become crises. I'll be keeping an eye on IBM.

      {"commentId":1997576,"threadId":"292223","contentId":"1572498","authorDomain":"jg2090"}
        Reply#4 - Wed Jun 18, 2008 5:15 PM EDT
        {"canLink":false,"threadId":"292223","isPrivate":false}
        Leave a Comment:
        You're in Easy Mode. If you prefer, you can use XHTML Mode instead.
        As a new user, you may notice a few temporary content restrictions. Click here for more info.
        {"threadId":"292223","contentId":"1572498"}
        Start TrackingStart Tracking
        Stop TrackingStop Tracking