Author: Vincent Diepeveen
Date: 14:50:21 07/06/03
Go up one level in this thread
On July 06, 2003 at 17:31:29, Tom Kerrigan wrote: >On July 06, 2003 at 09:07:24, Bo Persson wrote: > >>The argument was about Tom insisting that the 8080 couldn't have influenced the >>x86 instruction set, because its name didn't end in 86. It's not all in the >>name! > >No, no, no, no, no. Of COURSE the 8080 influenced the x86 instruction set. > >I'm just saying that the 8080 is not an x86. > >This is like saying the BMW 2002 was the first BMW 3-series. The BMW 2002 was >the predecessor to the 3-series, it influenced the design of the 3-series, it >was similar to the first 3-series in many ways, but it WAS NOT A 3-SERIES >because its model number DID NOT START WITH 3. > >Bob is hiding behind linguistic sloppyness to avoid admitting that he was wrong. Apart from quoting 80s truth. Where he is on Alzheimer drugs with respect to processors, in parallel search his memory stops around end of 80s when Schaeffer published his branching factor 10 thing called Aphid. Still calling that a 'good algorithm' in 2003 when branching factors are <= 3.0 Note that when you search well in RGCC around 1997 you'll find that Bob is saying in some threads there that i'm an idiot when i claim that the fullwidth branching factors of 4.5 to 5.0 are just too high to be used with nullmove, because he claims there that getting under 4.0 is impossible. His proof back then wasn't very impressive either back then, because it was using the Knuth formula which isn't taking into account hashtables nor nullmove. Somehow i get impression that the global 'store' in his brain stopped functioning back around 1990. Somewhere long before Chrilly's descpription of nullmove in ICCA which was around 1991. It is therefore common knowledge that at tournaments if some amateur programmer refers to hyatt when some pro's are in neighourbhood, that the pro's usually refer to hyatt as: "that man that still lives in the 80s". Perhaps we should remember him for that. The 80s with cray blitz at 100Mhz Cray processors which could do 29 instructions a clock and at 16 processors getting him the glorious 500k nps nodes a second with Cray Blitz, as it was so 'very well' vectorized (1.6Ghz K7 doing less than 3 instructions a clock gets crafty to 1MLN nps). Those were his glory days! In the meantime no dude here is posting much at RGCC anymore, with exception of Hyatt. It's revival of the 80s there, if you know what i mean. >-Tom 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.