Re: MAX Machine PLA

On 02/08/2016 20:39, Segher Boessenkool wrote:
> The 6566 was designed before the MAX was, afaik.  It clearly was designed
> for SRAM only fwiw.

Interesting, I heard that the decapped 6566 showed the dram logic hacked 
up to work with sram.
But it has a lower number, so it probably was done first.

>> The c64 had to
>> wait for them to ramp up production/yield (commodore failed to clone
>> dram when they tried).
> They did?  I never heard that story before, do you have any more details?

Failing to clone dram I think was from a Bil Herd talk. I've seen so 
many of them, but I think it was https://www.youtube.com/watch?v=SQ1IkpqIF1E

IIRC he doesn't give any details. Supposedly commodore decapped and 
copied a few chips over the years, which is ironic that ricoh did the 
same to the 6502 for the nes.

Waiting for dram I think came from commodore beyond the edge. The c64 
runs ram at 2mhz, the atari 8 bits (which was the only other home 
computer in the c64's class) ran it's ram at 1.79mhz and launched with 
8k. When atari started selling them with 48k, then they used 24 x 4116 
chips. In contrast, the c64 used 8 x 4164 chips for 64k.

To further highlight the state of the dram market at the time, Clive 
Sinclair was able to buy enough reject 4164's to  manufacture the 
spectrum. They binned them based on which half worked and then used 8 
for 32k, with a mother board trace cut depending on which bin they were 
using. Which is devious, but better than the xbox (where they bought 
defective ram and checked what speed it could run reliably at when 
booting, so your xbox could run slower than your friends).


       Message was sent through the cbm-hackers mailing list