Novag Citrine

Discussion of chess software programming and technical issues.

Moderators: hgm, Rebel, chrisw

Steve B
Posts: 3697
Joined: Tue Jul 31, 2007 4:26 pm

Re: Novag Citrine

Post by Steve B »

Hi Steven

you must really love your Citrine,,replying to a post nearly 6 months old
:P

Sadly there has only been a handful of posts about dedicated computers during all of that time so any and all post's about them are welcome
Sigh Regards
Steve
User avatar
mhull
Posts: 13447
Joined: Wed Mar 08, 2006 9:02 pm
Location: Dallas, Texas
Full name: Matthew Hull

Re: Novag Citrine

Post by mhull »

Steve B wrote:Hi Steven

you must really love your Citrine,,replying to a post nearly 6 months old
:P

Sadly there has only been a handful of posts about dedicated computers during all of that time so any and all post's about them are welcome
Sigh Regards
Steve
Steve and Steven,

If I were going to design a new dedicated computer, it seems like it would be interesting to find a suitable chip for which there was also a software emulator available for free. Then open up a chess programming competition that would be played out with perhaps thousands of games on a gang of hosting machines. It would be the only logical application of a uniform platform competition. The winner would become the official programmer of the unit for one year. Each year, open the competition again. Programmers would then have some incentive to improve their designs.

There could be variations on this theme where over time, the unit could load any program from past winners to new ones or the top three finishers.

On the hardware optimization side, I'd investigate whether the system could be designed to utilize SRAM only instead of SDRAM to reduce the number of clock cycles for a memory fetch.

I don't know how practical any of that would be.
Matthew Hull
wgarvin
Posts: 838
Joined: Thu Jul 05, 2007 5:03 pm
Location: British Columbia, Canada

Re: Novag Citrine

Post by wgarvin »

http://www.amazon.com/USB-Portable-Roll ... B000WX1FA4

A portable roll-up chess board with USB interface, for less than the cost of lunch. I wonder what it uses for sensors. Reed switches?
User avatar
sje
Posts: 4675
Joined: Mon Mar 13, 2006 7:43 pm

Re: Novag Citrine

Post by sje »

For more on the H8 microprocessor used in the Citrine (and some other dedicated units), see:

http://en.wikipedia.org/wiki/H8_Family

There are free H8 tools and an active developer community.
IanO
Posts: 496
Joined: Wed Mar 08, 2006 9:45 pm
Location: Portland, OR

Re: Novag Citrine

Post by IanO »

wgarvin wrote:http://www.amazon.com/USB-Portable-Roll ... B000WX1FA4

A portable roll-up chess board with USB interface, for less than the cost of lunch. I wonder what it uses for sensors. Reed switches?
My impression is that it is a pressure-sensory board. The protocol is extremely minimal, it transmits an algebraic coordinate when a square is pressed. I've also heard that the build quality is what you would expect for the price; don't expect it to last very long.
IanO
Posts: 496
Joined: Wed Mar 08, 2006 9:45 pm
Location: Portland, OR

Re: Novag Citrine

Post by IanO »

mhull wrote:
Steve B wrote:Hi Steven

you must really love your Citrine,,replying to a post nearly 6 months old
:P

Sadly there has only been a handful of posts about dedicated computers during all of that time so any and all post's about them are welcome
Sigh Regards
Steve
Steve and Steven,

If I were going to design a new dedicated computer, it seems like it would be interesting to find a suitable chip for which there was also a software emulator available for free. Then open up a chess programming competition that would be played out with perhaps thousands of games on a gang of hosting machines. It would be the only logical application of a uniform platform competition. The winner would become the official programmer of the unit for one year. Each year, open the competition again. Programmers would then have some incentive to improve their designs.

There could be variations on this theme where over time, the unit could load any program from past winners to new ones or the top three finishers.

On the hardware optimization side, I'd investigate whether the system could be designed to utilize SRAM only instead of SDRAM to reduce the number of clock cycles for a memory fetch.

I don't know how practical any of that would be.
My impression is that the Phoenix Revelation is just this computer. It uses a 500 MHz XScale processor and 32 MB of RAM. The XScale's ARM architecture is pretty much ideal for chess programming; every instruction can conditionally branch. It comes with Shredder, Rybka 2.x, Sjeng, Ruffian, and old freeware engines Toga and Fruit.

Additionally, Phoenix has developed both 68000 and 6502 emulators for their product, allowing all of Richard Lang's champion Mephisto programs to be emulated, as well as many of the best 6502 programs (Polgar, Rebell 5.0, MM V, and even the Super Expert C). This involved not only emulating the processor, but also the board and front panel interfaces of each of these dedicated computer programs.
Steve B
Posts: 3697
Joined: Tue Jul 31, 2007 4:26 pm

Re: Novag Citrine

Post by Steve B »

IanO wrote:
mhull wrote:
Steve B wrote:Hi Steven

you must really love your Citrine,,replying to a post nearly 6 months old
:P

Sadly there has only been a handful of posts about dedicated computers during all of that time so any and all post's about them are welcome
Sigh Regards
Steve
Steve and Steven,

If I were going to design a new dedicated computer, it seems like it would be interesting to find a suitable chip for which there was also a software emulator available for free. Then open up a chess programming competition that would be played out with perhaps thousands of games on a gang of hosting machines. It would be the only logical application of a uniform platform competition. The winner would become the official programmer of the unit for one year. Each year, open the competition again. Programmers would then have some incentive to improve their designs.

There could be variations on this theme where over time, the unit could load any program from past winners to new ones or the top three finishers.

On the hardware optimization side, I'd investigate whether the system could be designed to utilize SRAM only instead of SDRAM to reduce the number of clock cycles for a memory fetch.

I don't know how practical any of that would be.
My impression is that the Phoenix Revelation is just this computer. It uses a 500 MHz XScale processor and 32 MB of RAM. The XScale's ARM architecture is pretty much ideal for chess programming; every instruction can conditionally branch. It comes with Shredder, Rybka 2.x, Sjeng, Ruffian, and old freeware engines Toga and Fruit.

Additionally, Phoenix has developed both 68000 and 6502 emulators for their product, allowing all of Richard Lang's champion Mephisto programs to be emulated, as well as many of the best 6502 programs (Polgar, Rebell 5.0, MM V, and even the Super Expert C). This involved not only emulating the processor, but also the board and front panel interfaces of each of these dedicated computer programs.
Dont think the Rev emulates the Novag Super Expert (at least mine doesn't)but other then that
All correct
costs a pretty penny too

AND ..soon to be announced..
NEVER before seen since the first commercial release of dedicated chess computers...34 years ago...

I Can Say No More Regards
Steve
User avatar
sje
Posts: 4675
Joined: Mon Mar 13, 2006 7:43 pm

Re: Novag Citrine

Post by sje »

The Hitatchi/Renasys H8 microprocessor is fun to program at the assembly language level. Old pdp11 programmers will feel at home and those doing assembly coding for less orthogonal CPUs will feel a relief.

But having coded a Z80 chess program and all of the bitboard parts of an Motorola 68020 chess program, the most important thing I've leaned is to not do any more chess programs in assembler. Modern compilers are just too good at optimization.
User avatar
mhull
Posts: 13447
Joined: Wed Mar 08, 2006 9:02 pm
Location: Dallas, Texas
Full name: Matthew Hull

Re: Novag Citrine

Post by mhull »

sje wrote:The Hitatchi/Renasys H8 microprocessor is fun to program at the assembly language level. Old pdp11 programmers will feel at home and those doing assembly coding for less orthogonal CPUs will feel a relief.

But having coded a Z80 chess program and all of the bitboard parts of an Motorola 68020 chess program, the most important thing I've leaned is to not do any more chess programs in assembler. Modern compilers are just too good at optimization.
Looks like the SuperH line would be powerful enough to make a solid senior master program (at least). It might be able to run some older H8 programs too.
Matthew Hull
User avatar
jshriver
Posts: 1342
Joined: Wed Mar 08, 2006 9:41 pm
Location: Morgantown, WV, USA

Re: Novag Citrine

Post by jshriver »

mhull wrote:
Looks like the SuperH line would be powerful enough to make a solid senior master program (at least). It might be able to run some older H8 programs too.
Wasn't that the same CPU that was in the Dreamcast console? If so loved that processor.