Author: walter irvin
Date: 10:56:26 01/28/00
Go up one level in this thread
On January 27, 2000 at 10:42:34, Christopher R. Dorr wrote: >Deep Blue was a *Multi-Million* dollar machine. Significantly more than >"$100,000". > >It has a *ton* more computing power than you seem to think. And this power is >dispersed in ways vastly different from any single-processor architecture of >today. It's a completely different architecture than any single-processor PC. It i was not the one who thought it did not have great computing power .i know that it does .the other guy was suggesting that it had only slightly more computeing power than 1.5 ghz .and i was saying that surely it was faster than that . >is virtually impossible to make statements like you have made; it's kind of like >saying 'If a Chevy Malibu had the engines of the Space Shuttle, it would go >faster'. This is not true for a variety of reasons. The two are simply >incomparible on a superficial level, as is Deep Blue vs. today's typical micro >program. > >Christoph the statements i have been making is that given equal mhz pc's are as good or better .maybe deep blue is over rated??? deep blue's hardware is not over rated . if anything the main problem as i see it is deep blues hardware is vastly underated , plus the phycological impact a person would go through when playing such a beast .knowing it will miss nothing tacticaly .just knowing that db was searching 200,000,000 nps and how tuff pc's are at 250,000 nps along with db unpredictable play cause kasparov to start doubting himself .history of big chess matches show that once a player starts doubting himself he makes errors he normally would not make .so the main problem kasparov had was kasparov .db 1 and db 2 were not that different in elo .cm 6000 could have beaten kasparov by that final game , he was a beaten man . > >On January 27, 2000 at 07:37:24, walter irvin wrote: > >>On January 27, 2000 at 01:21:28, Jeremiah Penery wrote: >> >>>On January 26, 2000 at 18:00:29, walter irvin wrote: >>> >>>>Deep Blue has great results to its credit but it has had great hardware .the >>>>following programs at equal speeds are as good or better . >>>> >>>>1.cm 6000 >>>>2.fritz 6 >>>>3.junior 6 >>>>4.m-chess 8 >>>>5.shredder 4 >>>>6.tiger 12 >>>> >>>>at equal mhz programs like fritz 6 would get more nps im sure. >>> >>>Equal MHz, eh? Let's calculate the total MHz of DB's chess chips: 2.4 MHz/chip >>>* 480 chips = 1152 MHz. After you factor in that DB was only getting about 30% >>>efficiency, it was still doing 200M NPS, so that translates to 173611 NPS/MHz. >>>I don't think Fritz can calculate 173K NPS on a 1 MHz machine. Sorry. >>> >>>>they would search >>>>much deeper than deep blue. >>> >>>Fritz on an 1152 MHz machine might get 14 ply at 3 min/move. DB was getting the >>>same. Fritz has selective-search (null-move) errors. DB hasn't. DB extends >>>_way more_ than Fritz, so it sees way more than Fritz. >>> >>>>in computer vs computer matches speed kills , thats >>>>why on the ssdf , programs on faster hardware get inflated ratings . >>> >>>At equal MHz speeds, it's shown that DB is faster. Therefore, by your logic, it >>>would win. >>>At equal NPS, the programs are the same speed. Which one will win, then? >> >>i thought it had more computing power than that , surely it did .a 1 ghz kyrotec >>athlon is available ,you can get that less than $3000 . i thought deep blue was >>at least a $100,000 machine .
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.