Author: Robert Hyatt
Date: 11:19:00 01/26/04
Go up one level in this thread
On January 26, 2004 at 13:35:12, Christopher A. Morgan wrote: >On January 25, 2004 at 21:38:27, Robert Hyatt wrote: > >>On January 25, 2004 at 20:04:16, Rolf Tueschen wrote: >> >>>- - in a famous German forum the kids are on the streets and they shout: >>> >>>These old-fashioned Cray Blitz and Deep Blue monuments won't be "disqualified" >>>by their authors with actualized Elo numbers. >>> >>>Is that true? Would these legends lose badly against today's elite of >>>computerchess programs? >>> >>>I'm waiting! >>> >>>Rolf >> >> >>I don't believe _any_ of them would "lose badly". Any "super-program" from deep >>thought through Cray Blitz would be very tough opponents for today's programs. >>However, hardware is beginning to catch up. Someone just pointed out on a chess >>server last night that this quad opteron system I have is about the same speed >>as the Cray T90 I ran on in 1995, in terms of raw nodes per second (6-7M back >>then, 7-8M typically on the quad opteron). So it is now probable that Crafty >>could actually win a match from Cray Blitz on a T90 with 32 CPUs, assuming I use >>the quad opteron. My quad xeon 700 got ripped by the same machine a couple of >>years back, however, so it would still be dangerous. >> >>I can't say much about how it would compare to other commercial programs as I >>didn't run those tests with very little test time to play with the T90. >> >>The superiority of today's programs over the super-computers of 1995 are mainly >>mythical, IMHO. I suspect the games would be a _lot_ more interesting than some >>would believe. Of course, there is little chance to test such a hypothesis >>since most old programs are long-retired, and such hardware is not readily >>available today. > > >A couple of things seem clear. First, is that with today’s programs speed of >2-4 million nps is sufficient for the machines to play on a par with the super >GMs. Second is that the super GMs are very fearful of the machines. Their fear >is palpable. By now I believe we have had in excess of 25 man-machine games at >classical time controls against the very top players, and the machines are >playing at least in the 2750 ELO range. The top players can do mo more than >draw their short matches with the machines. Of course, not enough games, but >enough to know that in the opinion of the best human players, the ones who play >the machines in public matches, the machines are now at a par with the best >human players. I don’t believe with the older machines that this was even >claimed back then. I will remind you that deep thought did this _same_ thing against GM players. And that deep blue was 100X faster. And, of course, deep blue actually beat kasparov. My point is (here) not that deep blue is that much stronger than micros, since that is not easy to verify, but that _clearly_ todays micros are not "vastly superior" to deep blue either. Based on DT's and DB's record, if the micros were vastly superior, GM players would be losing 15:1 to them, but they are not. So far they are "breaking even". At best, you might argue that todays programs are in the same "class" as deep thought/blue" and make that argument pretty convincing based on past performances of micro programs. But to claim they are far better than DB has _no_ supporting evidence whatsoever, other than urban legend and myth... > IBM was so fearful of GK in 97 that they would not provide >GK with any of DB’s games, a far cry from today when the humans have hundreds of >the machine’s games to analyze before a match, and still can do no more than >draw the machines in a short match. IBM wasn't "fearful" of anything. They simply played by the rules as agreed on by both parties in the competition. More recent matches have had rules that were different. > >Given the above, it is hard for me to believe that the supercomputer >machines/programs of yesteryear could do any better against the machines of >today, despite their comparative slow speed, than today’s top players. > >Chris That's not the point being argued. The point being argued is whether or not today's micro programs are far superior to DB or CB or any other "big iron" program of the 1990's. The answer is, simply, "no". They might be pretty close for all I know. But _not_ "far superior"...
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.