Author: Robert Hyatt
Date: 19:49:48 07/21/00
Go up one level in this thread
On July 21, 2000 at 19:39:56, Chris Carson wrote: >Dr Hyatt questioned my qualifications (ASICS in particular) >I did not know I needed to supply a resume, but here is a brief >one. I have published this before for Dr Hyatt, last Feb when >he was debating Tom K. about an idea Tom had for creating a >program in hardware. Dr Hyatt had a fit and trashed Tom, when >I said Tom was right and I could do it at TI, Bob demanded to >know what my qualifications were. Look it up, it is in the >archives and some of you may remember. > >I hava a BS EECE 1984. I have a MS 1995. I worked for >Boeing 1985-1987 (systems engineer). I worked for LTV >1987-1989 (software engineer). I worked for Texas >Instruments 1989-15Jun2000 (Senior Engineer) in the ASIC's >(application specific integrated circuits) division and I have made >a lot of ASIC designs and delivered more chips than I can count >(but TI can) to external customers. Perhaps I am not an expert, >but I am not a novice. TI is paying for my PhD and I have completed >all courses and working on my disertation. This is a distraction. > >Let me be clear. I have left TI and I am now in private practice >as a Licensed Professional Counselor (LPC). My PhD is in the field >of psychology, I am making a career change. > >Dr Hyatt's logic is flawed in my opinion. DB hardware was at the 1995 >WCCC and finished tied for 3-5. I have given the links that prove that >from the tournament record and IBM (confirms that it was DB hardware in >1995). If you take that approach, I am not surprised you made a career change. There were two instantiations of the Deep Blue hardware. This is _well_ documented. DB1 was assembled just prior to the first Kasparov match. DB2 was assembled just prior to the second match. Prior to that there was _no_ deep blue hardware in existence. When IBM hired Hsu and the DT group from CMU, they had an internal name the chess machine contest. Deep Blue was the winner. Beyond that point they started calling the deep thought machine "deep blue prototype". If you look at any of the ACM tournament descriptions after Hsu went to IBM, you will find under the program name/description "Deep Blue Prototype - deep blue software running on the deep thought hardware." That is well documented. I have told you this several times. It has been found by others in many printed articles. I think the description is quite clear. They were fiddling with the search and evaluation tables. They used this 'fiddled version' but had no hardware to use except for the original deep thought.02 hardware built at CMU. Is that hard to understand? Of course, you can go to your local library, and look thru back issues of CACM and find the issues that cover each annual ACM and the less frequent WCCC events. You will find exactly what I said. Or you can email Hsu or Campbell. And ask. Rather than speculating or declaring as fact things you have only seen on a web site for the general public. Why not ask the _source_??? He answers email if it is polite and sensible. >So Fritz3 beat DB hardware (DT software if you like) and >Wchess drew it. DB hardware Has not played a public game against micro computers, since it didn't exist until 1996, well after Hong Kong. You can say it all you want. But it doesn't make it true. Just do your research.. >has a loosing record against P90 and >has no data for faster machines (the match of 36-4 was against P90 >and old sw, heck todays programs on only 450mhz have a pretty good >record see the SSDF against the P90 and old sw, let alone DJ 8x-700). > >DB made simple mistakes (ED, AMIR, URI) have all pointed this out. >Simple mistakes that are easily found by most programs on slower >machines (450mhz for example). > >DJ or Fritz on a 8x-700 is a fair match for 97DB. My opinion is >they would beat 96 DB and crush DT (based on games against humans >and comp vs comp results of Junior and Fritz on 450mhz machines). > >DB had a bizillion NPS. Ask Dr Hyatt, he will tell you all about >that. DB needed them due to a simple eval and still missed simple tactics >even with its bizillion NPS (or what ever DR Hyatt says). >A great accomplishment in 1997, twice as fast as the HW that lost >to Pritz3 and Wchess on P90's in 1995. DJ, Fritz, Rebel, Shredder, >and Hiarcs (perhaps more) have better evals in my opinion. > >DB 97 would never win 90% against Fritz and DJ on a 8x-700. > >Best Regards, >Chris Carson so speaks a true "expert" in the field??? Nice declarations of fact. With no weight at all behind them...
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.