WIERD EXPERIMENT SPOILER ALERT.
Somedya, fi I even get the time, I mnay take up this experiment once again.
the challenge: extra memory for al ower cost.
The answer: A hard drive.
NOTES:
Cartridge has two boards, each mapped to different areas of hardware. Top board, CPu ROM program, and sound sample section to the YM2610.
Bottom PCB: 32 bit wide graphics CHA lines with up to 256 megs of addressable space. FIX rom of 1 meg for text mode. SM1 sound rom of 1 meg.
Abstract: In the tradition of the Henemann's Sluggo unit, program ROM *CAN* be intercepted to use as a controller by using RAM as an emulator, a dual port RAM as it were. The idea being for an internal CPu with its own firmware for its functions. It would control 6 banks of RAM in the following order:
1. Internal cPu scratchpad and stack RAM
2. Neo Geo CPU ROM emulation RAM
3. Neo Geo Sample ROMspace emulation RAM
4. Neo Geo CHA emulation RAM
5. Neo Geo SM1 emulation RAM
6. Neo Geo SFIX emulation RAM
Basically, the cartridge CPU would load up each RAM bank with the proper data, switching on the fly as needed. It wouldnt be so much of a true coprocessor as a master arbitrator.
The first preliminary of this system used an Apple ][ GS with a 1 gig hard drive as the arbitrator. The first set of experiements were geared towards having multiple games on the same cartridge, as a memory test.
hard drive crash sucks

. had about a month of fun into that idea. Maybe some other time....who knows? Anyone want to hire me for about 2 years to finish up said project?
