Later we designed a 68000 program called Cyrus 68K, written by one of us (Mark Taylor), which evaluated and sorted all the moves from a node being expanded: we used these evaluations to determine the SXDEC for moves that were non-tactical. An obvious way to accomplish this is to assign the "best" (i.e., highest-scoring) non-tactical move a low SXDEC and to determine the SXDEC values for its non-tactical siblings on the basis of the difference in score among them.
... A second generation of the search extension algorithm was developed during the period 1985-1988. Many of the original ideas were used but we tried to eliminate certain obvious deficiencies and to make the intelligence in the program more sophisticated. We came up with a number of new ideas and tested them in a 68000-based program called Cyrus 68K. In general the results were rather encouraging, and we now feel that there is no longer a need to be shy about our work, hence the revised acronym for the algorithm and the renaming of the key variables to SEX and SEXDEC.
a chess program by Mark Taylor and advisor David Levy for the 68000 (68K) based family of microprocessors. Cyrus 68K, not to confused with Richard Lang's earlier program Cyrus, was written by Taylor from scratch in 1985. Richard Lang, like Mark Taylor affiliated with Levy's and O’Connell's company Intelligent Software, already focused on Psion at that time, had abandoned Intelligent Software, and was about to work for Hegener & Glaser on Mephisto.
Table of Contents
Photos & Games
SEX
Cyrus 68K was Intelligent Software's testbed for the SEX Algorithm, to apply fractional extensions and reductions, most notably already a kind of LMR, considering forced tactical moves such as checks, captures, mate threats determined by Mate at a Glance, as well as killer moves and early non-tactical moves and from a evaluated and sorted move list with a lower depth decrement than later non-tactical siblings.Excerpts from The SEX Algorithm in Computer Chess 1989 [4] :
Later we designed a 68000 program called Cyrus 68K, written by one of us (Mark Taylor), which evaluated and sorted all the moves from a node being expanded: we used these evaluations to determine the SXDEC for moves that were non-tactical. An obvious way to accomplish this is to assign the "best" (i.e., highest-scoring) non-tactical move a low SXDEC and to determine the SXDEC values for its non-tactical siblings on the basis of the difference in score among them.
...
A second generation of the search extension algorithm was developed during the period 1985-1988. Many of the original ideas were used but we tried to eliminate certain obvious deficiencies and to make the intelligence in the program more sophisticated. We came up with a number of new ideas and tested them in a 68000-based program called Cyrus 68K. In general the results were rather encouraging, and we now feel that there is no longer a need to be shy about our work, hence the revised acronym for the algorithm and the renaming of the key variables to SEX and SEXDEC.
The Sphinx
Cyrus 68K was commercially available as the Sphinx, a dedicated chess computer under the brand name CGX, and manufactured by Eric White's Newcrest Technology Ltd., Hong Kong [5] [6].See also
Publications
External Links
References
What links here?
Up one level