| Posted By
Litwr
on 2019-11-10
03:59:29
|  An article about 6502 and Commodore
Hi there
I dare to present my article about the heart of our best +4, 6502. It was published last year but, sorry, I forgot to send a message here, I was too busy. This emotional article is a part of my larger emotional article about first processors (z80, x86, ARM, 68k, ...). I hope that somebody can find it useful. Thanks.
https://litwr.livejournal.com/2773.html
https://litwr.livejournal.com/tag/cpu
|
|
Posted By
MikeZ
on 2019-11-10
07:47:29
 | Re: An article about 6502 and Commodore
Bravo! Excellent.
|
|
Posted By
Lavina
on 2019-11-10
09:25:02
 | Re: An article about 6502 and Commodore
Nice to see you buddy, welcome back!
|
|
Posted By
MMS
on 2019-11-10
16:03:06
 | Re: An article about 6502 and Commodore
Hi Litwr,
It was a very nice read! It is very detailed and contains a lot of interesting info I did not know beforhand.
So it is a great article, so the usual high quality material from you! 
You mentioned the 10MHz possibility of 6502. Actually the CMOS and HMOS was a much less energy efficient production technology, than even what indutry has in the late 90s.If you increase the frequency, the CPUs inner parts overheat (andI think this is what we see in 8501s' high ratio defect)
I am just curious, if the 6502 structure fully revealed (it is, I saw some Xray photoes), would it be possible to produce (clone) is with eg "ancient" 65nm process, that would make wattage very low for such a processor. It meand, with a proper cooling, it could go up to some hundred MHz without frying it. MOS is not existing any more.
Or the design (type of the gates) need to be modified/upgraded to able to keep up with the clock? Certainly I do not speak about prefetch and command pipelines, like in case of eZ80. (I am really not a HW expert)
|
|
Posted By
gerliczer
on 2019-11-11
00:16:32
| Re: An article about 6502 and Commodore
"would it be possible to produce (clone) is with eg "ancient" 65nm process"
That would be a world class financial train-wreck. No sense in it. It should be (and is already) done with some FPGA. There are most probably cheap ones that can run at a couple of dozen MHz. That would be an astronomic improvement.
|
|
Posted By
Litwr
on 2019-11-11
15:35:29
| Re: An article about 6502 and Commodore
Thanks for nice words. Bill Mensch several years ago clearly stated it is possible to make 6502 at 10 Ghz, I worked with hardware men recently and they confirmed this claim. Interestingly, Bill, in his oral history, said that 6502 would be used at 4 MHz in 1976 - it was really very fast. In 1995 he said that some chips were used at 10 Mhz in 1976 - 5 MIPS! - it was really shocking speed which could beat even the best mainframes of these days! However they did not have RAM that could match such fantastic performance.
|
|
Posted By
MMS
on 2019-11-11
16:03:54
| Re: An article about 6502 and Commodore
Thanks for the info.
I impressed by FPGAs (though I have no real knowledge on them), but it is still an emulation, and it's speed is limited by their internal clock. I know, because my IEEE2IEC interface is based on at ATMEGA8, an it's clock speed is not enough to catch up with JiffyDOS synch signal speeed and no way to solve it AFAIK.(told by its developer)
10GHz is really something. It would be nice to see an 6502 with nitrogen cooling, haha. SRAM needs no refresh, while it's speed is still far from demand. DDR4 is the word your are looking for Would be kind of funny to link the ancient 6502 with the latest PC RAM. (certainly it is not realistic, just kidding).
|
|
| |
Copyright © Plus/4 World Team, 2001-2024 |
