Author: Robert Hyatt
Date: 07:49:27 04/26/01
Go up one level in this thread
On April 26, 2001 at 10:21:38, Albert Silver wrote:
>On April 25, 2001 at 19:18:18, Vincent Diepeveen wrote:
>
>
>>
>>it shows regurarly 12(6) which would mean it has finished 18 ply fullwidht
>>according to Bob.
>>
>>That means even with a perfect branching factor
>>
>>40 (first ply) ==> not taken into account though
>>4 (factor 4 for overhead because of using big qsearch)
>>let's take an optimal branching factor for deep blue with SE in software
>>and also captures added.
>>
>>optimal branching factor fullwidth 4.5 (it's way more without any extensions
>>my b.f. is even bigger fullwidth).
>>added to that 0.5 for SE
>>added to that 1.0 for capture extensions
>>total b.f. = 6.0
>>
>>6.0 ^ 18 * 4 = 406239826673664 nodes needed.
>>
>>Ok let's forget even the factor 4 overhead for qsearch
>>then it's 101559956668416
>>
>>Bit big number?
>>
>>Oh well now let's take the number of nodes explored by deep blue.
>>200M nodes a second. Of that 20% was effective according to Hsu (see IEEE99
>>publication).
>>
>>So compared to a single cpu that would be 40M nodes a second.
>>Let's even be MILD here for Hsu and ignore the fact and just use 200M
>>a second, because Bob is going to complain.
>>
>>So even when taking the defense of DB into account the
>>b.f. was absolutely better as
>>
>>18th square from 200M x 3 minutes = 18th sq from 36B nodes.
>>that's about 3.86
>>
>>Now that's *impossible* without nullmove everywhere or very dubious
>>forward pruning. Note the example of Hsu he takes a 'typical 12 ply search'
>>as example.
>>
>>More likely this 8(6) says nothing and it's simply 8 plies in total from
>>which 6 in hardware. To solve many tricks he did the junior/belle trick
>>and he just extends in software about every tactical move he sees.
>>
>>That solves of course testsets bigtime.
>>
>>Also on the internet Campbell was recently searching during the
>>world championship FIDE with deep blue junior (i do not know how
>>many processors 30 or so?).
>>After many minutes of search he reported his variation was
>>a 12 ply search.
>
>I was there, and I happened to record what he said once I was made aware of it.
>Following the games with CA6 beta at the time (plug, plug) I was able to include
>these comments as a text comment in the notation. I have unfortunately only two
>such lines with a text note. Here is what I have with the positions being
>commented:
>
>[Event "Brain Games World Championship Match"]
>[Site "London, England"]
>[Date "2000.10"]
>[Round "10"]
>[White "Kramnik, Vladimir"]
>[Black "Kasparov, Gary"]
>[Result ""]
>[Eco "E54"]
>[Annotator ""]
>[Source ""]
>
>1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.e3 O-O 5.Bd3 d5 6.Nf3 c5 7.O-O cxd4
>8.exd4 dxc4 9.Bxc4 b6 10.Bg5 Bb7 11.Re1 Nbd7 12.Rc1 Rc8 13.Qb3
>Be7 14.Bxf6 Nxf6 ( 14...Bxf6 15.Nb5 Bxf3 16.Qxf3 a6 17.Nd6 Rc7 )
>15.Bxe6 fxe6 16.Qxe6+ Kh8 17.Qxe7 Bxf3 18.gxf3 Qxd4 19.Nb5 Qxb2
>20.Rxc8 Rxc8 21.Nd6 {rb8 nf7 kg8 nd8 h6 ne6 nh5 qa7, +.83, min depth 11, max
>depth 27 (Deep Blue Jr.)
>}
>Rb8 ( 21...Ra8 22.Nf7+ ( 22.Ne8 Ng8 ) Kg8 23.Nd8 h6 24.Ne6 Nh5
>25.Qb7 Re8 26.Qd7 Rb8 ) 22.Nf7+ ( 22.Qxa7 Rf8 23.Qe7 Ra8 24.Nf7+
>Kg8 25.Nd8 h6 26.Qf7+ Kh8 27.Ne6 Ng8 28.Re2 Qf6 29.Qb7 $20 )
>Kg8 {Nd8 h6 Ne6 Nh5 Qa7 Rc8 Qd7 Qc3 +.87, examining about 35 M pos/sec (DB Jr.)
>}
>23.Qe6 Rf8 ( 23...h5 24.Ng5+ Kh8 25.Qf5 Rf8 26.Qg6 Qb4 27.Re6
>h4 28.Nf7+ Kg8 29.Rxf6 Qe1+ 30.Kg2 h3+ 31.Kxh3 Qf1+ 32.Kh4 $20 )
>24.Nd8+ Kh8 25.Qe7
>{p-svidler(GM) kibitzes: I've never seen Garry playing so badly. First time in
>years he seems to be off-form in an important competition
>
>} *
>
>Here are the two FENS of the positions in question:
>
>[D]2r4k/p3Q1pp/1p1N1n2/8/8/5P2/Pq3P1P/4R1K1 b - -
>rb8 nf7 kg8 nd8 h6 ne6 nh5 qa7, +.83, min depth 11, max depth 27 (Deep Blue Jr.)
>
>
>[D]1r4k1/p3QNpp/1p3n2/8/8/5P2/Pq3P1P/4R1K1 w - -
>Nd8 h6 Ne6 Nh5 Qa7 Rc8 Qd7 Qc3 +.87, examining about 35 M pos/sec (DB Jr.)
>
>BTW, considering the irregularity of the notation (lowercase in the first and
>capital letters in the second), I think Murray was entering this by hand. Each
>comment came with the evaluation, and only after did he had the info on depth
>and speed. It wasn't all in a single line as I posted here.
>
>Before you go off on a bend, remember that this was still probably the software
>info being displayed on his screen as Bob mentioned, but I'll let you two argue
>this. I'm going to go get some more popcorn. :-)
>
>Last thing: he said it was something like 128 processors being used as I recall.
>
> Albert
>
I believe it was either 16 or 32. I don't think a single-cpu SP can drive
128 of them. This is a "balancing act". As you add chess processors, the
SP has to search less deeply so that it can give each chess processor something
to do quickly enough...
I am vague on the 16 or 32, but I did chat with him during the games a few
times... I'm pretty sure it was one or the other. the 32M nodes per second
would suggest 16 processors... 128 at 2-2.4M nodes per processor would be quite
a beast..
>
>>
>>Now that was like after a lot of time. And with the huge b.f. of
>>Deep Blue at those depths (as SE and recapture extensions
>>go take their toll on bigger depths in software), it's very likely
>>that Deep Blue only searched 1 ply deeper in a bit shorter time!
>>
>>So there is a source confirming this indirectly!
>>
>>He told this to channel 211 too, so there are more witnesses if they
>>still remember the questions i asked regarding search depth!
>>
>>Best regards,
>>Vincent
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.