I wrote a go-playing program for a Transputer array in the 1980s. It won a 9x9 
championship, but not 19x19.
For 19x19 the search space is so large that some intelligence wins out over 
brute force.
Transputers were not a good design, in that they had no virtual memory, and 
inter-processor communication was quite slow.
Reliability was also a problem.

Jack Lang


-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of alain Baeckeroot
Sent: 13 January 2007 14:51
To: [EMAIL PROTECTED]; computer-go
Subject: Re: [computer-go] Can Go be solved???... PLEASE help!

Le samedi 13 janvier 2007 15:06, Don Dailey a écrit :

The first "Connection-Machine" CM1 (from Thinking Machine Inc) was
65 536 transputer connected on a 12d hypercube (one transputer at each corner)

Itw was quite hard to program, but i think it could be a very good hardware
for a strong go program :) Sadly it is now in museum.

Alain
_______________________________________________
computer-go mailing list
[email protected]
http://www.computer-go.org/mailman/listinfo/computer-go/

_______________________________________________
computer-go mailing list
[email protected]
http://www.computer-go.org/mailman/listinfo/computer-go/

Reply via email to