Author: Tom Kerrigan
Date: 18:05:37 06/18/02
Go up one level in this thread
On June 18, 2002 at 20:22:41, Robert Henry Durrett wrote: >On June 18, 2002 at 19:02:45, Tom Kerrigan wrote: >>So a common misconception seems to be that 64-bit chips would be twice as fast >>as 32-bit chips. >I doubt that any computer-oriented people would think that way. My wife >wouldn't either but that's because she neither knows nor cares what a "bit" is. >[And, why should she?] I hope my bulletin didn't seem to suggest that I thought >that way. No, you didn't specifically say "twice as fast" but I did get the impression from your post that you thought 64-bit chips are inherently better than 32-bit chips. >use those "64 bit chips" for. If you are doing simple addition, that's one >thing. On the other hand, a few of the many innovative and creative chess >engine inventors/developers here at CCC might have found OTHER ways to take >advantage of the 64-bit feature. It might, even, be useful! There are many chess programs that are based on bitboards--64 bit integers where each bit corresponds to a square and is used to store an aspect of the board, e.g., which squares are attacked by a certain piece. This seems like an optimal use of the datapath width to me. Crafty is one such program, but as I said in my last post, Crafty does not seem to benefit [significantly?] from the wider datapath. >>A chip's design is MUCH more important than how wide its ALU is. >I guess you are referring to that dubious/nebulous "software bandwidth" concept >here? I haven't heard of the "software bandwidth" concept... what I was referring to were other, more important aspects of a chip's design, e.g., how many execution units it has, how deep its pipeline is, how good its branch predictor is, how mnay registers it has, etc. If somebody gave me a choice between a 64-bit datapath and, say, an on-die memory controller, I'd take the latter in a heartbeat. >>The Itanium 1 is a dog. Everybody says the Itanium 2 will be fast, but according >>to what Intel itself has disclosed about performance, it's not going to be >>faster than a fast Pentium. Plus, it'll be really expensive. Personally, I think >>the Itanium's design is stupid and am waiting for it to disappear. >What matters is nodes per second, not clock rate. That is very important! True, and the Itanium loses on both counts. :) The 2.533GHz Pentium 4 is 193% faster than the 800MHz Itanium 1, according to SPECint. Intel says the Itanium 2 increases per-clock performance by 50% and it clocks 25% higher, so the 2.533GHz P4 will still be 56% faster. -Tom
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.