Mephisto RISC1 & RISC2 Strength Question

[Dave C]

I was reviewing dedicated chess computer ratings recently on the wiki site (http://www.schach-computer.info/wiki/in ... -Elo-Liste) and on Nick's website (http://www.spacious-mind.com/html/ratin ... ments.html).

I noticed that the Mephisto RISC1 modules (1992, Gideon 3.0) are rated a few points higher than the RISC2 modules (1994, Gideon 3.1). Does anyone have a theory why the newer modules with the updated and presumably stronger Gideon program are rated slightly weaker?

Somewhat confused...regards,
Dave

[Sargon1972]

Hello Dave,

Well depends on the games results in games vs other computers , the 3.1 is a better program ,but the 14mhz is a "killer" it is in fact to slow this CPU
Look at the same elo list and see the Chessmachine Gideon 3.0 and 3.1 there is a small difference of 24 elo ,but here comes the difference because this Programs run at 32mhz and with the style Aggressive and they are the same programs as Risc 1 (3.0) and Risc 2(3.1) and there should be small mistakes in version 3.0 that is why 3.1 is released without these small errors soon after
The Risc 1 ore 2 can also put on aggressive,but again here the cpu is to slow for this setting
so that they are rated a few points of each other is just results vs other computers
Nice thing to know Ed Schroder did leave Mephisto because he did not get the hardware he wanted ,a year later when Ed Schroder had joined Tasc he became WC and Mephisto wanted to but that Program from Ed because they had seen that they hade make a mistake by letting him go


Kr,Hans

[Dave C]

Although 14 mhz is clearly not as strong as the TASC 32 mhz, your description indicates that 14 mhz is too slow. Are you saying 14 mhz is inadequate to run the Gideon program effectively or is it just not as strong as a 32 mhz chess computer?

Thank you Hans.

BR,
Dave

[Sargon1972]

Hello Dave,

Well to slow for the strongest setting yes ,but think that the modules 14mhz Risc 1 where sold by Mephisto for 1000 Euro ,so what would be 32mhz cost....to much i think
Also double speed gain normal each chess computer +-100 elo ,with setting Aggressive it is a bit more for the 32mhz version
Nevertheless the Risc module set is very wanted under collectors and it can play a good game of chess

Kr,Hans

[steffen]

" it can play a good game of chess"

--> these praising words from a strong chess player as Hans is (Hans is nearly an FM).
Respect!

[Steve B]

Speaking about the Mephisto Risc Modules....

bidding hot n heavy for this Risc 2 Set

http://www.microsofttranslator.com/bv.a ... Swxp9W5BKN

Mucho Dinero Regards
Steve

[Lars Sandin]

I was reviewing dedicated chess computer ratings recently on the wiki site (http://www.schach-computer.info/wiki/in ... -Elo-Liste) and on Nick's website (http://www.spacious-mind.com/html/ratin ... ments.html).

I noticed that the Mephisto RISC1 modules (1992, Gideon 3.0) are rated a few points higher than the RISC2 modules (1994, Gideon 3.1). Does anyone have a theory why the newer modules with the updated and presumably stronger Gideon program are rated slightly weaker?

Somewhat confused...regards,
Dave

We in the SSDF tested both versions and for us, the 3.1 was rated higher:

Code: Select all

230 Meph RISC 2   1 MB                      2127   61   -66   125   39%  2207
237 Mephisto RISC 1 MB ARM 2  14 MHz        2103   19   -19  1356   55%  2068

210 Chess Machine 30-32 MHz Schr?der 3.1    2208   28   -27   646   63%  2120
220 Chess Machine 30 MHz Schr?der 3.0       2175   65   -55   168   73%  2003

The Risc 2 has very few games compared to the Risc 1MB, and unluckily I can't do anything about it as I only have the Risc 1MB (Gideon 3.0) in my Collection. I have both versions for the Chessmachine 32 MHz-card though.

Best regards