[Beowulf] MIPS/Watt data
Many of your questions may have already been answered in earlier discussions or in the FAQ. The search results page will indicate current discussions as well as past list serves, articles, and papers.
Bari Ari bari at onelabs.comWed Jun 2 07:57:12 PDT 2004
- Previous message: [Beowulf] MIPS/Watt data
- Next message: [Beowulf] MIPS/Watt data
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Jim Lux wrote: > Just for grins.. > > A company (Faraday Technology) has announced an implementation of a ARM > V4 core that they claim runs at 500 MHz at 0.3 Watts. They claim this > is good for about 675 Dhrystone MIPS (but, of course, since Dhrystone is > integer and small, it's probably running entirely out of the 32k on-chip > cache). That's 2250 MIPS/watt, which is quite impressive (of course, > it's on an ancient benchmark that is probably irrelevant for most > applications) > > If we compare Dhry mips to Bogomips (who's to say if we're not within an > order of magnitude).. A 3.4 GHz P4 turned in 6700 Bogomips. That was > probably around 100W (total guess), for 67 Bogomips/watt. ARM, Mips and SH all beat the pants off x86 in the MIPS/watt and also in the FLOPS/watt department. Nobody seems to be interested in clusters built with these architectures. We started design on a 1TFLOP cluster a few years back based on a 64/128 bit MIPS cpu by Toshiba (similar to the emotion engine in the Playstation2). It would have fit into a single 19" rack and consumed around only 6,000 watts. -Bari
- Previous message: [Beowulf] MIPS/Watt data
- Next message: [Beowulf] MIPS/Watt data
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the Beowulf mailing list
