• news
  • WEDNESDAY SEPTEMBER 20 2006 4:00 PM

Feel the Power of GFLOPS

You all know about distributed resource computing, where a bunch of people run software that uses their computer's idle cycles to search for alien communication or to fold protiens.

Now a Californian company has released software that allows users to use the power of their graphics card processor, at least for floating point operations and when they're not running games.

Compared to graphics processors, main system processors have developed much slower in providing an increase in floating point performance. For example, a Pentium 4 3 GHz processor was estimated to reach about 6 GFLOPs (billion operations per second), while Woodcrest, the server variant of Intel's Core 2 Duo processor, is currently believed to top out at around 24 GFLOPs; according to Intel, a four-processor dual-core Itanium 2 system recently reached 45 GFLOPs.

At least in the floating point discipline, graphics chips are way ahead of the game: ATI recently said that its current high-end X1950 XTX processor brings in 375 GFLOPs, in dual-graphics mode even up to 750 GFLOPs - the equivalent of 31 Xeon 5100 processors. Nvidia's Geforce 7950 GX2 dual-GPU card is rated at 384 GFLOPs and Ageia's physics processor at 96 GFLOPs.


The drawback is that they charge $2000 per computer node, so I'm hoping the Open Source community can work out something similar. I fancy having a super computer in my home.

 
Comments
turin

turin

Denver, CO
October 2003

SEP 20, 2006 04:27 PM

goodness, that's a lot of flops.

Spaceboy

Spaceboy

Dallas, TX
October 2004

SEP 20, 2006 05:25 PM

turin said:
goodness, that's a lot of flops.



I know, It's like a computer version of Al Gore

xxxxxxx

xxxxxxx

Canada
September 2005

SEP 20, 2006 06:54 PM

It's strange that that hasn't been done before now if it is possible and offers that much potential with floating point operations. I do hope something in the sub-$2000 range comes along.

BOINC!

AceT

AceT

Portland, OR
April 2004

SEP 21, 2006 05:41 PM

It's about time. Since the Folding@Home project began people have been saying this kind of thing would be considerably more effective if it used the GPU instead of or in addition to the CPU. They've touted performance increases up to 4000% if this was done.