Author: Robert Hyatt
Date: 12:52:43 04/03/02
Go up one level in this thread
On April 03, 2002 at 13:23:57, Keith Evans wrote: >On April 03, 2002 at 11:44:45, Robert Hyatt wrote: > >>On April 03, 2002 at 04:26:37, Tom Kerrigan wrote: >> >> >>> >>>>Deep Thought's evaluation was based directly on Belle's... Deep Thought >>>>was simply a "belle on a single chip" where belle was a huge batch of FPGA's >>> > >Actually I don't think that this is quite correct. Deep Thought was not Belle on >a chip - at least not according to Hsu's thesis. Deep Thought was the ChipTest >move generator (part of Belle on a chip), plus some flavor of sequencer, plus >some programmable logic etcetera to handle the evaluation. As far as I can tell >the Deep Thought evaluation described in Hsu's thesis was simpler than the Belle >evaluation. > >I think that you said that ChipTest did some part of the search in a different >post, but this is not correct. It was just a move generator. Search was handled >by microcode running on an external sequencer. I hope I didn't say much about chiptest, because it didn't last very long. But deep thought was definitely "belle on a chip" according to the write-up by Hsu in the book "Computers, chess and cognition". Farther into the article, Hsu says "the main difference between the processors is the enhanced hardware evaluation function." (where 'the processors' is referring to the chiptest machine vs the deep thought chip.) > >Maybe there were subsequent versions of Deep Thought which integrated more >functionality on a chip - but I didn't think that this happened until the era of >Deep Blue. Please correct me if I'm wrong. So far as I know, deep blue was simply an evolutionary update of the original deep thought single-chip search engine... It added features (evaluation and others) deemed important enough to add. DB2 added even more features and addressed some repetition detection issues along with a significantly enlarged evaluation circuit... > >>>I don't think FPGAs had been invented at the time. >>> >>>-Tom >> >> >>They've been around forever in various forms. Look up Ken's paper on >>Belle. Or his paper "An FPGA based move generator for the game of Chess" >>or any of several other publications including the second edition of >>"Chess skill in man and machine." >> >>Belle used them in 1980, for certain. I don't know how long they had been >>out by then... > >I think that Belle must have used some form of FPLA or PAL which is a little >different than an FPGA. I remember using some of the early Xilinx FPGA parts in >the early 1990's are they were quite primitive as that time. (As far as I know: >FPLA - programmable AND and OR terms, PAL - programmable AND terms, CPLD - >usually multiple PALS plus some programmable routing, FPGA - totally different >structure) > >Here's a quote from Xilinx: > >"Founded in 1984 and headquartered in San Jose, California, Xilinx invented the >field programmable gate array (FPGA) and fulfills more than half of the world >demand for these devices today." > >Regards, >Keith No idea about the above usage of FPGA. Ken used that acronym in a paper he wrote prior to 1984...
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.