Author: Andrew Dados
Date: 22:09:27 01/28/05
Go up one level in this thread
On January 28, 2005 at 15:30:10, Vincent Diepeveen wrote: >On January 28, 2005 at 15:14:57, Steve B wrote: > >>>comeon tournament versions in past were 10000+ dollar a piece. >> >>the only dedicated computer alive today that can perhaps bring in $10,000 is >>Tasc 40(Version 2.5 King) with SB 30: >> >>http://pg.photos.yahoo.com/ph/cpaa2@sbcglobal.net/detail?.dir=3353&.dnm=69c4.jpg&.src=ph >> >>and perhaps Prototype of Boris Handroid: >>http://pg.photos.yahoo.com/ph/cpaa2@sbcglobal.net/detail?.dir=3353&.dnm=7a38.jpg&.src=ph >> >>Steve > >Not really, if i would make a huge wooden box, same size like tournament >machines those days. Some fans inside and say a cell type processor inside which >runs massively parallel eating little power, cost $20000, many would pay for it. > >(oh for those who want to know why, a cell processor can deliver 250 gflops, as >opposed to a normal processor like 2.4Ghz A64 which is fastest single cpu >processor at the moment delivers 4.8 gflops) How much of marketing speech did IBM and Sony put into that we'll have to see yet. For intro to cell processor see http://www.blachford.info/computer/Cells/Cell0.html For lazy: basically 1 cell processor contains a 'processor unit' which is something like G5 (running supposedly at 4.6 Ghz hmm I have yet to see THAT in 2005) and 8 APUs From article: "Each Cell contains 8 APUs. An APU is a self contained vector processor which acts independently from the others. They contain 128 X 128 bit registers, there are also 4 floating point units capable of 32 GigaFlops and 4 Integer units capable of 32 GOPS (Billions of Operations per Second). The APUs also include a small 128 Kilobyte local memory instead of a cache" And also: "The APUs are vector [Vector] (or SIMD) processors, that is they do multiple operations simultaneously with a single instruction. Vector computing has been used in supercomputers since the 1970s and modern CPUs have media accelerators (e.g. SSE, AltiVec) which work on the same principle. Each APU appears to be capable of 4 X 32 bit operations per cycle, (8 if you count multiply-adds). In order to work, the programs run will need to be "vectorised", this can be done in many application areas such as video, audio, 3D graphics and many scientific areas." Good luck utilizing 8 FP apus and writing your own memory cashe code to run your favorite chess app :) It is like having to program 8 Nvidia GPUs now. (a modern graphic processor on your graphic card deliver more gflops then your pentium/AMD and yet noone is writing chess programs on them) (note: chess programs with exception of some parts like move generator and parts of evaluation routine is really hard to 'vectorize' especially in FP) So if you don't spend 3 years coding/debugging then cell == G5. -Andrew- > >Actually i'd love to buy a small cluster with diep at it. But i don't have the >money for it. At tournament machines you never make profit actually. It just >costs money. > >Vincent
This page took 0.04 seconds to execute
Last modified: Thu, 07 Jul 11 08:48:38 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.