Author: Dave Gomboc
Date: 23:19:34 10/19/02
Go up one level in this thread
On October 19, 2002 at 20:32:45, Jeremiah Penery wrote: >On October 19, 2002 at 19:50:02, Uri Blass wrote: > >>On October 19, 2002 at 19:29:15, Jeremiah Penery wrote: >> >>>On October 19, 2002 at 17:03:51, Dave Gomboc wrote: >>> >>>>On October 19, 2002 at 14:45:19, Steve Lim wrote: >>>> >>>>>SJLIM: Alot of programmers on CCC have asked me to ask you this.. for >>>>>clarification.. >>>>>SJLIM: Please explain search depths for the notations 4(5) and clarify earlier >>>>>comments about 12(6). This may include indicating what is "normal full width" >>>>>searching, extensions, quiesence search, or other types of searching DB2 >>>>>utilized, and which was done in software versus in the hardware chess chips. >>>>>SJLIM: Also, what types of pruning were used. This topic has generated enourmous >>>>>discussion on CCC. >>>>>CrazyBird: 4(5)means the same thing. 5-ply maximum hardware depth, although it >>>>>is obviously impossible in this case. >>>>>CrazyBird: since the brute force depth is 4. >>>>>CrazyBird: i can't really go into the details of the hardware pruning. it is >>>>>related to method of analogy pruning, or rather a basterized form of it. >>>>>CrazyBird: limitation in the contract with ibm. >>>>>SJLIM: Can this be answered? - Does 12(6) mean the 6 is included _in_ the 12, or >>>>>in addition to the 12? >>>>>CrazyBird: 6 is part of 12, but the hardware can search less than 6, that is the >>>>>software horizon may be more than 6 plies. >>>>>CrazyBird: and of course, the selective depth can be arbitrarily deep, well, no >>>>>more than 8 times brute force. >>>>>CrazyBird: argh, the q search. it is in hardware. both sides are allowed checks >>>>>in quiescence search. max is 8, i think. >>>> >>>>Clearly, Vincent's interpretation is supported here. >>>> >>>>I'm still sure they were searching much deeper than 12 ply in important lines >>>>though! >>> >>>Strange that he said 4(5) is an impossible case, when at least half of their >>>searches had that depth reported. Also, I'm not sure how, if the second number >>>is a maximum depth, that it could be included in the first number, when the >>>second number is variable (not to mention that the second number is often bigger >>>than the first). It implies that the first number (software depth, not counting >>>extensions/pruning) is variable also (which Hsu said), but I'm not sure how that >>>would work. >>> >>>>>CrazyBird: 6 is part of 12, but the hardware can search less than 6, that is >>the software horizon may be more than 6 plies. >>> >>>Say the software searched 7 plies (or more) - how do we know that the hardware >>>didn't also search 6 plies beyond that (we just know it didn't search MORE than >>>6 plies). >>> >>>Maybe I'm being a bit obtuse, can anyone try to explain this to me? >> >>If the software searched 7 plies then it told the hardware to search 5 plies and >>not 6 plies >>see the following: >> >>"2.Does 12 means that the depth of the software in deeper blue was less >>than 12 plies(12-x when x is the depth of the hardware that is not constant)? >>CrazyBird: yes, the software "brute force" depth is always less." >> >>Hsu agreed that 12 means 12-x in the software when x is the depth of the >>hardware. > >It seems like in 12(6), software depth is anywhere between 6 and 11, and the >hardware makes up the rest of the 12. It makes sense, though it is a little >strange, and he could (should) have very easily made it clear before. But I >still don't know how to make sense of 4(5), when Hsu says it was impossible, yet >it happened in most of their searches. If the number is simply wrong, then what >assurance is there that all other the other depth numbers are correct? > >Hsu could have just said "12 is software+hardware depth", but instead he says >"12 in terms of brute force" before, and then in response to your question he >said "the *software* 'brute force' depth is always less." Why does he use the >term brute force in both of those statements, which gives the impression that he >uses the term to describe only the software search? Of course, if his comments >today were correct, then the hardware should also be counted in the 'brute >force' depth, but he doesn't make it very clear in his statements. I think it is because, going down one variation from root to tip, there is software brute-force, then software selective, then hardware brute-force (with analogy pruning), then hardware selective.
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.