|
It sounds possible. The 8080 on which the very first IBM PC was based, clocked just north of 4 MHz. This was late 1970s technology.
|
|
|
|
|
The ZX-81 used the same processor as the ZX-80: a Z80A @ 3.25Mz. (or the NEC µPD780C-1 equivalent, depending on the manufacturer stock).
This was not as fast as it sounds in comparison to the BBC and VIC-20 as the whole thing was designed to be as cheap as possible and was horribly slow.
Bad command or file name. Bad, bad command! Sit! Stay! Staaaay...
|
|
|
|
|
OriginalGriff wrote: This was not as fast as it sounds in comparison to the BBC and VIC-20
Why not though? Both 8 bit processors presumably with similar instruction sets (I'm familiar with the 6502, not the Z80), so I would have thought that clock speed would be a good indicator of performance.
Regards,
Rob Philpott.
|
|
|
|
|
Processor is only one part of a system: think how slow a Windows PC gets if you have too little RAM, for example.
The ZX81 had only four chips IIRC - processor, RAM (1K), ROM (8K), and a programmable logic chip to tie it all together and provide the chip selects etc.
The BBC had a lot of hardware to do things the Sinclair had to do in software - so more instructions need to do the same task.
Bad command or file name. Bad, bad command! Sit! Stay! Staaaay...
|
|
|
|
|
No. The 6502 actually was a snail when compared to a Z80 at the same frequency. Then again, in such contests algorithmic optimizations often had more effect and triggered the next round when the opponent did the same optimizations on 'his' processor.
It has been decades since I wrote code for a 6502, but from my deepest memories comes 'zero page addressing' as one of the causes of the 6502's slowness.
The language is JavaScript. that of Mordor, which I will not utter here
This is Javascript. If you put big wheels and a racing stripe on a golf cart, it's still a f***ing golf cart.
"I don't know, extraterrestrial?"
"You mean like from space?"
"No, from Canada."
|
|
|
|
|
CDP1802 wrote: but from my deepest memories comes 'zero page addressing' as one of the causes of the 6502's slowness
I might be a bit old now, but wasn't the point with Zero Page Addressing that you only needed two bytes instead of three when addressing and thereby making it 30% faster if you could fit the core of the code into the first page?
Personally I remember swearing over just having three registers.
|
|
|
|
|
Yes, I think you are right. It has been a while, but not so long I used a C compiler for the 6502 to write some stuff for the 8 bit Ataris and a C64. The author of the C compiler also did not think that the 6502 was ideal for C. Nor any other 8 bit processor, I think. Still, it's real fun to chase some oldschool sprites around the screen with C. 
The language is JavaScript. that of Mordor, which I will not utter here
This is Javascript. If you put big wheels and a racing stripe on a golf cart, it's still a f***ing golf cart.
"I don't know, extraterrestrial?"
"You mean like from space?"
"No, from Canada."
|
|
|
|
|
Also, the Motorola 6502 and Intel 8080 used the clocks differently - the 6502 was '2 cycle' and the 8080 was '4 cycle' for the same instruction processing unit, so 1 a 4 MHz 8080 was *roughly* equivalent to a 2 MHz 6502.
According to my calculations, I should be able to retire about 5 years after I die.
|
|
|
|
|
Other sources seems to be reporting the same CPU speed.
I'd rather be phishing!
|
|
|
|
|
Yeah that's a little hard to believe. But if you go to this[^] vintage site they say the same thing.
New version: WinHeist Version 2.1.0
My goal in life is to have a psychiatric disorder named after me.
I'm currently unsupervised, I know it freaks me out too but the possibilities are endless.
|
|
|
|
|
I think he had a 6.5MHz clock, the 3.25 being derived from it for the processor.
|
|
|
|
|
Yes, that bit makes sense. I think computers of the day tended to have the cpu on one clock cycle, the graphics chip on the next. It's just the speed that surprises me.
Regards,
Rob Philpott.
|
|
|
|
|
Just checked and it is also documented in the English wiki. My source was a French site for Vintage computing technology.
|
|
|
|
|
The 6502 was a RISC CPU running at max 1MHz and you can think an average of 2-3 cycles per op.
The Z80A was more a CISC CPU running at max 4MHz.
Let's say the Z80 was a bit more sophisticated and easy to program than the 6502, which was much like a bunch of raw flip-flops...
Just kidding: loved 6502!
|
|
|
|
|
Mario Vernari wrote: The 6502 was a RISC CPU
Urgh! You've opened up a can of worms there. I fall into the CISC camp.
Regards,
Rob Philpott.
|
|
|
|
|
Aaaargh!
The 6502 was no RISC CPU! Look at it's register set and addressing mode. Does not really look reduced, right?
Could it be that you mean the CDP1802, which probably really was the first RISC CPU on a single chip. The predecessor, the CDP1801, still was made up of two chips. And by the way, it also was one of the earliest CMOS CPUs.
Both things made it unpopular. CMOS was considered to be slow and RISC (for lack of the proper words) was usually just called 'weird'.
The language is JavaScript. that of Mordor, which I will not utter here
This is Javascript. If you put big wheels and a racing stripe on a golf cart, it's still a f***ing golf cart.
"I don't know, extraterrestrial?"
"You mean like from space?"
"No, from Canada."
|
|
|
|
|
Well, you're right from a theoretical viewpoint. From a practical one, 6502 and other cores such as the Motorola 6805 and big bros were RISC when compared to Z80 and Intel's.
The internal architecture and the BUS was even much simpler than Z80.
Thanks for point it out, though...it's roughly 30 years I don't use them...
|
|
|
|
|
Who told you that? RISC does not really mean that a processor is inferior and only has a spartan instruction set.
A RISC processor usually has a large set of general purpose registers, CISC processors have a small set of registers, each dedicated to a specific purpose. In this respect the 6502 is definitely CISC.
A CISC processor usually has many different address modes and several variants of each instruction to support different combination of source parameter and destination parameter addressing. In this resprct the 6502 again is CISC. A RISC processor does not have all those addressing modes and emulates them over the general purpose registers. The 'reduced instruction set' means only that there are not a hundred variants of each instruction to support all the addressing modes.
The language is JavaScript. that of Mordor, which I will not utter here
This is Javascript. If you put big wheels and a racing stripe on a golf cart, it's still a f***ing golf cart.
"I don't know, extraterrestrial?"
"You mean like from space?"
"No, from Canada."
|
|
|
|
|
Nothing about inferiority: I was just pointing the relative simplicity of the logic of a 6502 than a modern processor. At the time, the 6502 was a myth!
Well, about RISC/CISC definition, looks like you're right: don't want to discuss that.
I did NOT study computer science, I'm an electronic engineer, thus I'm not sure about many definitions and theory behind computers.
I programmed enough 6502, 6805, AVR, but also x86 and a very minimal Z80: all that using assembler. What I remember is the huge diversity in the typical instruction ability, as well as the relative cost for doing something, up to the dramatic hardware (e.g. BUS) difference between CPUs.
Whereas 6502, 6805 (it's a MCU family, just consider the core) were very simple as architecture and hardware, had a small yet "trivial" set of ops even fast, the counterpart Z80, x86 and so were much more complex and powerful. I remember the 386 opcodes book which was huge and terribly complex.
So far:
- 6502, 6805 were a reduced (small and simple) instruction set? Yes: easy to check.
- Z80, x86 were a complex (vast and powerful) instruction set? Yes: easy to check as well.
To close, maybe both the 6502 and the 6805 should be classified as CISC, but I can't see them much different than a RISC. I believe they're borderline.
The "Indirect addressing" section of Wikipedia explains pretty well that:
http://en.wikipedia.org/wiki/MOS_Technology_6502[^]
Thank you anyway for clarify the differences.
|
|
|
|
|
Have you looked at an ARM processor at all? 
And they describe that as "RISC" ! 
Bad command or file name. Bad, bad command! Sit! Stay! Staaaay...
|
|
|
|
|
In 1987 I would have bought a Acorn Archimedes, but I never even got to see one. It was the first computer withn an ARM processor. I love RISC. Still, my first computer still has a CDP1802, which was first sold in 1976, 11 years earlier.
The language is JavaScript. that of Mordor, which I will not utter here
This is Javascript. If you put big wheels and a racing stripe on a golf cart, it's still a f***ing golf cart.
"I don't know, extraterrestrial?"
"You mean like from space?"
"No, from Canada."
|
|
|
|
|
Dividing a frequency by 2 is not really hard. Or by any other power of two. Still, it only matters what goes into the CPUs clock input, not what frequency it is derived from.
The language is JavaScript. that of Mordor, which I will not utter here
This is Javascript. If you put big wheels and a racing stripe on a golf cart, it's still a f***ing golf cart.
"I don't know, extraterrestrial?"
"You mean like from space?"
"No, from Canada."
|
|
|
|
|
Why? My little CDP1802 is from 1976 and its maximum clock frequency is at 3.2 MHz. The holy wars back then were about processors and, despite being the enemy, the Z80 (if I remember right, that's the CPU in a ZX-81) was available in versions of up to 4 MHz at that time.
Weak memory, but I think it was the Z80B that went up to 4 MHz.
Edit: The Z80A went up to 4 MHz. The Z80B went up to 6MHz. Just looked it up.
The language is JavaScript. that of Mordor, which I will not utter here
This is Javascript. If you put big wheels and a racing stripe on a golf cart, it's still a f***ing golf cart.
"I don't know, extraterrestrial?"
"You mean like from space?"
"No, from Canada."
modified 2-Mar-15 11:13am.
|
|
|
|
|
And the last version of a Z80 I used (one of the Z180 variants with 1MB address space via built in MMU) was running at 32MHz.
It was a very nice little processor, and I was quite sad to abandon it in favour of the ARM family - but I needed the processing power, and the Z80 was just too slow.
Bad command or file name. Bad, bad command! Sit! Stay! Staaaay...
|
|
|
|
|
How many MHz does a piece of wire have?
I wanna be a eunuchs developer! Pass me a bread knife!
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
