I realize on a site called 6502.org yall are probably a little biased, but in your opinion, which of the cpus from the 80s was the most elegant/most powerful/just plain best?

That's really three different questions, so I just answered with my favorite, rather than getting caught up in the "usefulness per expenditure" argument. Elegance is a very subjective area, and has the hardware and software facets.

Mike B.

It's widely held that the 6809 is very good, and the 6309 is even better. Of course, my own favourite is the 6502.

(Edit: just spotted that this is a poll. What's 65k - is that meant to be 68k? Because of course the 68k is another kind of thing altogether. Or is it meant to be 65816?? I've never seen it referred to as 65k.)

Eheheheh... uh, my mistake. I mixed up 6502 and 68k and somehow came up with the 65k as being a seperate 8bit cpu... XD

In some ways, it might be like arguing over what's the best favorite color. Most of us here do have the 6502/65c02/65816 as our favorite, but it sometimes comes down to personal reasons, including sometimes what we got into first or had had the most exposure to, rather than that one is necessarily best.

Most processors have more than enough power to do a wide range of jobs; so then what makes one "best" may be factors of availability in the form desired (the 6502 scores poorly here if you want a family of 6502-based microcontrollers with flash- or (E)EPROM-based program memory that can be programmed on the workbench), understandability (6502 scores very high here), how well documented it is (the 6502 comes out on top), availability (6502 is not available at very many distributors like a lot of other processors are, but it is being made today, with no end in sight), interrupt performance (6502 is probably the best of all the 8-bitters), etc..

The following is from the "Intro: Why a 6502?" section of the 6502 primer:

• very simple bus structure. You don't have to be a computer engineer to design and build with it. It's a great place to start.
  
  
• an assembly language that makes it easy to envision solutions, also making it a good place to start assembly-language programming. Michael Barry said on the 6502.org forum, "Yeah, programming the 6502 is like a guilty pleasure for my brain. It seems to know what I'm trying to say, and it just does it without complaining about the atrocious way I said it."
  
  
• deceptively high ratio of computing power to complexity, and it out-performs many of its contemproraries which on the surface appeared better
  
  
• possibly the best interrupt performance of any 8-bit processor (See my interrupts article)
  
  
• all of zero page is basically 256 bytes of processor registers (It's neither RISC nor CISC, but ARA, "addressable-register architecture.")
  
  
• way more than enough stack space for many languages (according to tests, and contrary to common misconception), accessible to the programmer
  
  
• still being made and available today, with no end in sight, and in hobbyist-friendly packages (DIP and PLCC) that fit in thru-hole sockets, including wire-wrap
  
  
• very well supported, and long-term support is expected to be better than that of many other processors that are popular today. It is probably also the most documented processor.

Companies find it attractive for new embedded controller designs because of the small number of gates, small licensing fee, and, according to Windbond, easy development; and as a result, the 6502 is being made today in huge volumes (over a hundred million units per year). They're rather invisible though, controlling processes in automotive, industrial, toy, appliance, and even life-support equipment, inside custom ICs. (You won't have to pay any licensing fees, but my point is that the low fee for processor manufacturers is one reason the 6502 will remain in wide usage for a long time to come.)

The podcast here is an Aug 2015 interview with Bill Mensch, pres. of WDC who holds the IP for the 65c02 and 65816, regarding these processors and comparing them to ARM, 68000, x86, 6800, 6501, etc., and his business model and his goals. He obviously has a very clear vision, and he is accomplishing what he wants. Still. Today. In the March 2015 podcast interview with WDC's David Cramer, WDC's VP of business development, he says there are hundreds of different products being made today with 65xx processors in them.

Nostalgia might be a reason for some to use the 6502, but for me it's not. Even 30 years ago however a Z80 had to have a clock speed of 4MHz to keep up with a 1MHz 6502; and Jack Crenshaw, an embedded-systems engineer who wrote regularly in Embedded Systems Programming magazine said in the 9/98 issue that he still couldn't figure out why, benchmark after benchmark, the 6502 could outperform the Z80 which had more and bigger registers, a seemingly more powerful instruction set, and ran at higher clock rates. Since that time, the 6502's clock speed has increased by an order of magnitude (actually the fastest ones run over 200MHz, but the ones available to the hobbyist are limited to about 20MHz) and the instruction set has improved (especially on the 65816). Assemblers and compilers have improved, making it easier to develop and maintain software.

Sophie Wilson, chief architect of the ARM processor, said, "an 8MHz 32016 [National's 32-bit processor —GW] was completely trounced in performance terms by a 4MHz 6502." The 65816 even outperformed the 68000 and 8086 in the Sieve of Eratosthenes benchmark.

So yes, the 6502 / 65c02 / 65816 has a lot of pluses, including many which don't immediately meet the eye. Those are not necessarily what make a processor best for a particular person or job though.

The 6809 and 6309 Ed mentioned were upgrades to the 6800. The 65CE02 was one big upgrade to the 6502 that was in production for a while, but the primary upgrade, which is still in production, with no end in sight, is the 65816, which I think is a better upgrade. "65K" refers generally to the family, 6502, 65c02, 65816, 65832, and others, some of which were never in full-scale production.

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

I don't usually participate in polls but I'll jump in here real quick to wave the flag for the MSP430 family -- so, so simple, but powerful, and a pure delight to code for in assembly. (Wikipedia page here.)

PS to Garth -- I guess you know that saying the 6309 is an upgrade to the 6800 is like saying the '386 is an upgrade to the 8080! Both statements are true, but a lot gets lost in the translation...

_________________
In 1988 my 65C02 got six new registers and 44 new full-speed instructions!
https://laughtonelectronics.com/Arcana/ ... mmary.html

I love the 6502 for its simplicity. It is probably the best microprocessor for beginners who want to be able to build their own microcomputer and understand how the entire thing works (as Garth said, it is very well documented and there are tons of example projects that newbies can draw inspiration from).

That being said, for situations in which more demanding software is required / desired (running the linux or minix kernels for example) something more powerful such as the 65816 or 68k is more appropriate.

_________________
- Billy

It's nice to see that we have various ideas of the ideal micro, despite being great fans of the 6502.

From the web:

although a later commentator disagrees:


TI's own note SLAA024 says

In that the 6809 was a step between the 6800 and the 6309? The 6309 has a great instruction set but still had only a 16-bit address bus and never got past 5MHz.

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

I don't disagree, and TBH I'm not even taking the discussion seriously. All I meant was, the 6809 (and, even more so, the 6309) is miles ahead of the 6800. The word "upgrade" hardly does them justice.

_________________
In 1988 my 65C02 got six new registers and 44 new full-speed instructions!
https://laughtonelectronics.com/Arcana/ ... mmary.html

I spent about 1.5 years trying to get into Z80 assembler in the early 80s. It was quite capable, but I never liked how different the registers were. I found myself often wanting another address register, but HL was already in use... or wanting another counter, but BC was doing something, etc. It being my first assembly language, it was just inexperience on my part, but in retrospect it's telling that, rather than my first being my favorite... it's instead the one I have the least nostalgic fondness for.

The mid 80s were spent on the 6502, which was (perhaps counter intuitively) freeing in its lack of registers. I've heard people describe the 6502 as having 256 registers (i.e. zero page ARE registers). That's one way to look at things, but I'm not that generous. The 6502 has a single accumulator and a couple of index registers... not 256 registers. But it's incredibly freeing to be released from "register stress". You're never wondering what register to use because there (a) aren't many registers and (b) the registers there are are only useful for short fragments. Variables live in memory and registers act (temporarily) on variables. It's a mindset shift of sorts. I found that I could be very productive with the 6502. I had a blast.

The late 80s were spent with the 68000, and it was like stepping into another level of computing. Hardware multiply and divide, multi-position and variable shifts, multi-byte data, memory poi.... well... if you've used the 68k I don't have to rant on, and if you haven't then by now you never will, so I won't bore you. Suffice it to say as I moved onto more powerful machines the 68000 was up to the task.

So, for me, 68k > 6502 > Z80 at this point (say, 1988).

For quite random reasons I stumbled upon a 6809 over the summer of '89 and I spent some time with it. I actually loved it. It felt like a step between the 6502 and 68000.

So, 68k > 6809 > 6502 > Z80 (by the end of 1989).

In the 90s I worked on commercial VME-based 68k systems, and later systems that used 68360 microcontrollers. By the end of the 90s I was working on embedded PowerPC 603e systems. PowerPC, with it's virtual memory and other modern operating-system features was another eye-opener. As was the move from CISC to RISC.

Later I had the opportunity to use both ARM and MIPS. PowerPC, ARM and MIPS are all similar. ARM is rather elegant, and it's doing quite well.

My final score for the 80s 8-bit CPUs, then: 6809 > 6502 > Z80.

The 6502, Z80 and 8086 are all from the 70s, so I vote "Other". From the 80s, I think the Transputer series of chips are worth an honourable mention.

I second this - and it's more or the less the reason most of us are here.

Thanks to sark02 for that concrete timeline of experience! I did just a very little with the 68k, and I liked it. But my next step, although it came much later, kind of dates from 1984:

The 70s passed me by - consisting only of the Atari VCS (2600) that was way too expensive for my family. The UK home computing boom in the 80s was my first introduction to programming, with the ZX Spectrum. My older sister's boss was really into computers, and she would always bring home his old copies of Your Computer, Personal Computer World, BYTE, and others. This is from where I learned BASIC, and then (albeit poorly) Z80 - from a magazine series.

I self-taught myself 6502 assembler as soon as I got my hands on an Atari 800XL, and barely touched its BASIC which (I later found out) was pretty bad.

So 6502 is my spiritual "first", and so it stands as my favorite.

ARM, though... well... ARM is extraordinary fun. I haven't touched it since '04, so I'm out of touch with the modern cores, but I'm sure I'll fall into a job that lets me work with it again in the next few years. Being an embedded software guy in the networking/systems space, it's inevitable (although x86 is becoming awfully popular).

BTW, I voted 'Other' for the 6809, because I do think it's superior, but my understanding is that it was quite a bit more expensive, too... which is why it only found its way into niche products (e.g. Dragon 32 in the UK, Tandy TRS 80 in the US). 6502 was _everywhere_, so it was a much more transferable skill.

Well, if we're picking from any 8/16-bit microprocessor, I've got to give my vote to the LSI-11 series - the PDP-11 architecture is pretty much the perfect balance of easy low-level accessibility and higher-end flexibility/functionality. The TMS-9900 series is also quite good in this regard (not surprising, since it's pretty much wearing the 11 on its sleeve.)

That said, though, if you want to go with something simpler, the 6502 is the most comfortable, usable minimalistic design I've ever encountered - just enough features and functionality to do pretty much anything you need, and simple enough that it's easy to learn how to put the individual puzzle pieces together to accomplish what you want. Plus, being as cycle-efficient as it is certainly doesn't hurt.