Computer graphics during the 8 bit era

[AndyC]

repeat: the processor IS making FULL use of MEMORY time when 6502 cycle = RAM cycle
in this case the memory can't do better, can't support any extra access in half-cycles
is it so bloody hard to understand?

The memory in a 6502 system has to be capable of responding within 1/2 of a CPU cycle and so has to support two reads per CPU cycle. This is an unavoidable consequence of the CPU design. So 6502 cycle = RAM cycle is just something you can't have.

This, by the way, is exactly why the 6502 systems typically have a clock rate slower than equivalent Z80 systems.

[1024MAK]

Not entirely true AndyC, you can have slower RAM and as long as the set-up and hold times are okay for the RAM to respond, and for the 6502 to be able to read data from the RAM. So it’s not always as simple as the 6502 having half a clock cycle for RAM access. With slower RAM, it may require more than 50% of the PHI2 clock for the set-up times to be met.

The determination is between the timing of the 6502, the access time of the RAM and the set-up and hold times.

But in the Acorn BBC Micro models A, B and B+, the Master series and the Commodore 64 (and possibly other systems) the system was carefully designed so that the RAM was fast enough for the video system to access the RAM in between the 6502 (or equivalent) microprocessor RAM accesses. In the C64, the processor clock was 1MHz (or thereabouts). In the Acorn machines, it was 2MHz. Hence the DRAM chips in these machines had to be capable of 4MHz operation (that’s partly the reason these machines were more expensive).

Mark

[Lethargeek]

Now you are disagreeing with others on here and with me. It could just be a misunderstanding. So maybe you would like to go into greater detail to explain what you mean?

i mean such a pic alone tells us nothing, you need to place it against a memory cycle(s)

The memory in a 6502 system has to be capable of responding within 1/2 of a CPU cycle and so has to support two reads per CPU cycle. This is an unavoidable consequence of the CPU design.

no, it hasn't, and in Ataris it wasn't, and in c128 2MHz mode it wasn't

So 6502 cycle = RAM cycle is just something you can't have.

i can have it any time i want with c128 FAST mode

This, by the way, is exactly why the 6502 systems typically have a clock rate slower than equivalent Z80 systems.

no, it has nothing to do with it, it was all about memory cycle to fetch time ratio

[RMartins]

We can all keep saying:
- No
- No, unless xxx
- Yes, with a BIG IF and several conditions

And you can keep repeating:
-Yes, unconditionally

It won't take us anywhere, and some less proper words have been used, which is not a good thing to do.
Retro computer is supposed to be fun and friendly.

So I have an interesting retro project for you Lethargeek.

The best way to remove any doubts, is for you ( who says it's possible unconditionally), to retro fit your idea into such a retro machine and then document how it works, with data analyser outputs and osciloscope traces, to prove your idea as valid.

I, for one, will be waiting anxiously to be proven wrong and that all the pre-requisits I and others mentioned were not required to make it work.

By the way, this could be a very interesting project.

[Lethargeek]

We can all keep saying:
- No
- No, unless xxx
- Yes, with a BIG IF and several conditions

And you can keep repeating:
-Yes, unconditionally

no, i keep repeating "yes, because it all was done before" (and for the most part in the c64 even)

It won't take us anywhere, and some less proper words have been used, which is not a good thing to do.

yes, pretending not to notice my examples and keeping saying noes will take us nowhere

Retro computer is supposed to be fun and friendly.

and forum messages supposed to be read and understood
and examples supposed to not to be ignored
there won't be any friendly fun otherwise

So I have an interesting retro project for you Lethargeek.

The best way to remove any doubts, is for you ( who says it's possible unconditionally), to retro fit your idea into such a retro machine and then document how it works, with data analyser outputs and osciloscope traces, to prove your idea as valid.

I, for one, will be waiting anxiously to be proven wrong and that all the pre-requisits I and others mentioned were not required to make it work.

By the way, this could be a very interesting project.

even if i was a hardware man, why should i waste my time for some obstinate naysayers amusement on a project that brings nothing new to the discussion, as there is perfectly working example of Atari 8bit since 1979, and most of the "pre-requisits" are already present even in c64/c128

[1024MAK]

But this is a Sinclair forum centred on the ZX Spectrum machines. We may not have a full detailed understanding of the Atari eight bit machines or a full detailed understanding of the Commodore 64 or 128 machines.

I have however used, repaired and build microprocessor systems using the 6502 microprocessor.

Mark

[Lethargeek]

But this is a Sinclair forum centred on the ZX Spectrum machines. We may not have a full detailed understanding of the Atari eight bit machines or a full detailed understanding of the Commodore 64 or 128 machines.

then why not reading a bit more about them before all these noes?
especially after these things were mentioned here

no deep dive is necessary to understand the Atari doesn't have fixed video memory reading pattern
https://en.wikipedia.org/wiki/ANTIC

[1024MAK]

Because there is only so much hobby time available to me...

Mark

[RMartins]

...
even if i was a hardware man, why should i waste my time for some obstinate naysayers amusement on a project that brings nothing new to the discussion, as there is perfectly working example of Atari 8bit since 1979, and most of the "pre-requisits" are already present even in c64/c128

Don't take me wrong, but sometimes, it looks like you just like to argue.

From my understanding, the debate is not about if it's possible or not.
Because It's possible under the right conditions and setups ( your mention of existing machines that do it is evidence of that ).

However, what has been argued by several, as I see it, is that you can't just change that particular machine to work like that, without requiring a completely different setup, reworking the entire architecture of the machine.

So I hope at least on this last bit we can agree on.

[Lethargeek]

Don't take me wrong, but sometimes, it looks like you just like to argue.

i just don't like when someone tells i'm wrong when i'm right

From my understanding, the debate is not about if it's possible or not.
Because It's possible under the right conditions and setups ( your mention of existing machines that do it is evidence of that ).

initially it was about WHY c64's CPU was so slow
and then how it was possible to make a faster 6502 system
if 3/4 of the VIC space weren't spent on sprite functionality
with the same-speed RAM chips

[1024MAK]

I think it was you that called the 64 slow

Mark

[Lethargeek]

I think it was you that called the 64 slow

huh?

The Commodore 64 used a version of the 6502 and ran it at a unremarkable speed, so it could not shift as much data around in a given time compared to some of the competitor machines.

[1024MAK]

Yeah, unremarkable, as in typical for the time. By competitor, I was meaning the far more expensive Acorn BBC Micro.

Mark

[Lethargeek]

Yeah, unremarkable, as in typical for the time. By competitor, I was meaning the far more expensive Acorn BBC Micro.

Maybe 'typical' for much simpler and cheaper systems, but for all the other complex hardware it was definitely lower end. The BBC Micro/Master was fastest, but there were several 6502-based systems in between. Also i think main competitor for the c64 worldwide was NES console (1.66-1.79MHz) rather than any proper home computer.

[Matt_B]

The Apple II ran at around 1MHz, but I suppose that was a much older machine. It wasn't until 1984 that they made a 2MHz model of it though, so there's a fair bit overlap with the C64.

The Oric also ran at 1MHz and has a clever, if somewhat limiting when it comes to making games on it, video system based around Teletext graphics and inline attributes. It might not have amounted to much in the long run but was the closest machine to competing with the Spectrum when it came to value for money in 1983.