chess software for dedicated chess computers

[bturner]

I'm at an early feasibility stage of an initiative to build a range of chess computers. The initial model would be similar to the Novag Citrine, a large table top computer. One aspect of the feasibility is determining pricing for all the component parts of the unit (packaging, software, electronics, marketing, etc).

I'd need to emulate the code onto a chip, and then design a motherboard and other electronics. As you probably know, the Citrine uses a H8/300 chip which only has 512KB of RAM. So I'm not looking for windows code - I need low level code that is good for a chip.

If anyone can assist with locating code, I'd greatly appreciate it. No point in reinventing the wheel - there's got to be lots out there.

[Kempelen]

I'm at an early feasibility stage of an initiative to build a range of chess computers. The initial model would be similar to the Novag Citrine, a large table top computer. One aspect of the feasibility is determining pricing for all the component parts of the unit (packaging, software, electronics, marketing, etc).

I'd need to emulate the code onto a chip, and then design a motherboard and other electronics. As you probably know, the Citrine uses a H8/300 chip which only has 512KB of RAM. So I'm not looking for windows code - I need low level code that is good for a chip.

If anyone can assist with locating code, I'd greatly appreciate it. No point in reinventing the wheel - there's got to be lots out there.

I am not absoluty sure, but I thing there are tons of chips that can be programmed in C. A simple search in google show a few entries about it.

[bturner]

I must be searching for the wrong thing - any suggestions as to a search phrase?

[hgm]

Let me get this straight: you want to have assembly code for some obscure chip to play chess (or do what?), and suppose that it would be available as public domain code on the internet???

Seems to me your only chance is to look for a C compiler for this chip, and compile open-source engines with it yourself. Problem could be that most top engines would run like cr*p (or not at all) when only 512KB of memory is available. E.g magic bitboard, which almost everyone uses nowadays, already requires 800KB for the Rook table. And then we are not even talking about a trasposition (hash) table yet.

Best chances for an engine with a small memory footprint would be a mailbox engine like Fruit. You might have to work on shrinking the hash-table demand, though. (Other memory-intensive tables, like the material table and pawn hash, can be switched off.) You will have less total RAM than even the most obsolete Celeron CPU will have cache...

[jshriver]

If I may recommend wait till the Raspberry Pi machine comes out hopefully in another month. It'll be $25-$35 and be a powerhouse 700mhz cpu, 128megs ram. Runs linux so porting any linux engine over would be easy. Plus it's the size of a credit card with HDMI out, ethernet, USB, and i/o pins which you could use for i2c with whatever sensors system you use.

Not that I haven't been thinking of doing this myself

[bturner]

Thanks Muller and Shriver for your comments.

Yes - I was hoping that code may be available from the web - it's viewed as sub-optimal these days and 100's of programs have been developed. I'm not, however, expecting it to be free. But to reinvent the wheel would cost $(tens of thousands). Something that's been used in previous dedicated computers could be reused in a different product.

One thing to bear in mind is that strength isn't the only wow factor in a chess computer. There's a niche demand for chess computers for people that would never use the huge processing capability of a big PC based chess program. OK this product wouldn't win any strength awards, but it would make a decent product.

The Citrine is a good example of a dedicated computer - but it's reliability and quality is a problem. It's just a Renesas H8/3687 chip which has 56k of program memory and 3k of RAM with CPU speed 20 MHz. Code for this type of chess computer must be available somewhere.

[smatovic]

The Citrine is a good example of a dedicated computer - but it's reliability and quality is a problem. It's just a Renesas H8/3687 chip which has 56k of program memory and 3k of RAM with CPU speed 20 MHz. Code for this type of chess computer must be available somewhere.

I also suggest to use some x86 Chip with an C Compiler and feed the machine with some of the best free Engines available.

[sje]

The Citrine is a good example of a dedicated computer - but it's reliability and quality is a problem. It's just a Renesas H8/3687 chip which has 56k of program memory and 3k of RAM with CPU speed 20 MHz. Code for this type of chess computer must be available somewhere.

I have a Citrine and I've found it to be totally reliable after much use and over an extended period of time. I've had it play many hundreds of games against my own program over a serial connection with no problems, Well, except for my cat who likes to jump on the board and bat the pieces.

The Renasas chip is essentially a DEC pdp-11 CPU pumped up with extra byte oriented instructions and extended memory. There is a complete SDK available, and any moderately sized program that can live in a mostly 16 bit environment can work.