65C816 vs 68000

[Dr Jefyll]

sark02 wrote:

why should I spend time with this new one if I'm just pounding nails and I have the perfect hammer already?

GARTHWILSON wrote:

Well put.

Yes, very well put. sark02 really hit the nail on the head.

[ttlworks]

Looked up 68020 versus CPU32 this weekend.

Compared to 68020, CPU32 misses the following instructions:
BFxxx: BFCHG, BFCLR, BFEXTS, BFEXTU, BFFFO, BFINS, BFSET, BFTST
CALLM, RTM, CAS, CAS2
cpxxx: cpBcc, cpDBcc, cpGEN, cpRESTORE, cpSAVE, cpScc, cpTRAPcc
PACK, UNPK
And there are no floating point instructions.

For CPU32, if Bit 6 of the addressing extention word is 1, Bit 2..1 have to be 0.
IMHO this rules out some indirect memory addressing modes:
([d]) ([d],d) ([Reg]) ([Reg,d]) ([Reg],d) ([Reg,d],d)

Also, stack format is different.

68020 has 6 different stack formats:
$0 and $1 (8 Bytes), $2 (12 Bytes) and $9 (20 Bytes) if interrupt meets coprocessor instruction,
$A (32 Bytes) and $B (92 Bytes) for bus error.

CPU32 has $0 and $2, plus $C (24 Bytes) for bus error.

;---

68020\CPU32 doesn't just push PC and status register on the stack,
but in some cases a few more Bytes about the internal CPU status etc.

Besides that, 68k has 8 data registers and 8 address registers 32 Bit, that's another 64 Bytes.
If there is a FPU present, that's another 8 registers 80 Bit each (80 Bytes) ?

6502 pushes 3 Bytes on stack if an interrupt hits (PC and status register),
pushing A,X,Y on stack is three more Bytes, and that's it.

65816 in native mode pushes 4 Bytes on stack if an interrupts hits:
PBR, PC, status register.
DBR (8 Bit) plus A,X,Y,direct_register (16 Bit) would be 9 more Bytes.

So we could say, that 68k clearly wasn't invented for a fast real_time response to interrupts.

When it comes to interrupt response\latency, 6502 feels more like a Kolibri.
65816 still feels like a Kolibri, although like a "slightly obese" Kolibri in native mode.

Despite MOVEM.L, 68k has the amenity\daintiness of an elephant when it comes to responding to interrupts...

[cbmeeks]

ttlworks wrote:

...It's an interesting question why there were more 68000 computers than there were 65816 computers.

Apple IIGS had a graphical user interface, so the reason can't be that the 65816 might be "too weak"
compared to the 68000 when it comes to "mouse pushing" while having colored windows on the screen.

The IIgs never sold that well. The 65C816 was a logical choice for the IIgs because Apple wanted to maintain backwards compatibility with the rest of the II series of computers.

Also, the IIgs was purposely kept slow by Jobs (as reported) because he didn't want it to compete with the Macintosh. Imagine a IIgs running at 8 MHz. Why would you ever want a Mac at that point?

[ttlworks]

cbmeeks wrote:

Also, the IIgs was purposely kept slow by Jobs (as reported) because he didn't want it to compete with the Macintosh.

Now that's interesting: Commodore didn't want the C65 (3.54MHz 65CE02 core) to compete with the Amiga.
But instead of just slowing down the C65, C65 went axed.

Hmm... there is a joke, that in hardware design only 5% of the decisions really are technology related.
The rest is tied to politics, logistics, existing tool chains... and dogma.

[cbmeeks]

ttlworks wrote:

Now that's interesting: Commodore didn't want the C65 (3.54MHz 65CE02 core) to compete with the Amiga.
But instead of just slowing down the C65, C65 went axed.

I believe the IIgs was allowed to live because, at the time, the Macintosh was still on unproven ground. The IIe was still the "bread-n-butter" of Apple and the Lisa and Apple III were failures. So I believe, from what I've read, the IIgs was a logical next step.

Also, initially, I think the IIgs actually sold pretty well. It just never sold the millions like the IIe did. So, I wouldn't call it a failure. I also believe it did a better job being backwards compatible with the IIe than the C128 did for the C64. The IIgs was a quantum leap from the IIe but the C128 wasn't all that much better than a C64. This was helped by the "Mega II" chip on the IIgs.

The IIgs really was an underrated computer. 4096 colors. 32 voice synthesizer. Fully 16 bit with 8 bit IIe mode.

From what I understand, Woz and many "Woz followers" created the IIgs. They were just crippled by Jobs. Woz designed computers that he wanted...which, turns out, I want too. Woz-designed computers appealed to the engineers in us.

ttlworks wrote:

Hmm... there is a joke, that in hardware design only 5% of the decisions really are technology related.
The rest is tied to politics, logistics, existing tool chains... and dogma.

Unfortunately, the same can be said for software.


**EDIT**

Now that I think about it, the 65C816 was underrated too. Other than the SNES and the IIgs, I can't think of anything it was used in that sold in large quantities. Makes me wonder why it's still around. There must be some use for it or they wouldn't keep making them.

I guess when the 68020 - 68060 came out, the '816 just couldn't compete. Especially when paired with FPU's. Of course, I image an '816 could utilize an FPU too.

This really makes me wonder what kind of computer could be designed around a 14MHz '816 these days.

[Tor]

cbmeeks wrote:

From what I understand, Woz and many "Woz followers" created the IIgs.

Over on woz.org Woz says he didn't, but that he may have helped inspire it (he was working on another advanced II project which was eventually cancelled, at the time).

Quote:

Woz designed computers that he wanted...which, turns out, I want too. Woz designed computers appealed to the engineers in us.

Agreed. The IIc didn't sell very well (no slots), the IIgs sold better (slots were back), and Woz forced the slots into the first II to start with (Jobs didn't want them). Without the slots I don't think the II had succeeded as it did. Looking back, everything I did professionally with the Apple II depended on the slots, which we used for various add-on boards, home-made and otherwise.

[Tor]

cbmeeks wrote:

This really makes me wonder what kind of computer could be designed around a 14MHz '816 these days.

Hmm. BDD's POC2? Unix-like filesystem, multitasking (IIRC), and other advanced features.

[cbmeeks]

Tor wrote:

Over on woz.org Woz says he didn't, but that he may have helped inspire it (he was working on another advanced II project which was eventually cancelled, at the time).

Ah. I was thinking he had more input because there was a limited edition IIgs signed by Woz. But like you said, he may have been more inspiration than hands-on designing.

Tor wrote:

cbmeeks wrote:

This really makes me wonder what kind of computer could be designed around a 14MHz '816 these days.

Hmm. BDD's POC2? Unix-like filesystem, multitasking (IIRC), and other advanced features.

That's right. I forgot about those. There's also the SBC3 too, right? Do you have a link to BDD's version?

[Tor]

cbmeeks wrote:

That's right. I forgot about those. There's also the SBC3 too, right? Do you have a link to BDD's version?

He hasn't yet put it up on his website I believe (the POC V1 is there though), so it'll have to be the following thread: viewtopic.php?f=4&t=1688

[ttlworks]

cbmeeks wrote:

This really makes me wonder what kind of computer could be designed around a 14MHz '816 these days.

Speaking of politics and logistics, that's a tough question.

We only have one source for the 65816, without WDC one would have to buy NOS chips from "obscure ebay sources".
Considering nowadays mice, keyboards and USB sticks, the computer needs some USB 2.0 master ports.
Maybe a CPLD/FPGA would be needed for video generation, analog VGA might be getting replaced by HDMI at some point.

And so on, and so on...
I'd like to see something like that... someday... but this probably should go into a different thread.

IMHO building a "stand_alone" 65816 computer with the features required would have to be a team effort,
but when discussing about this in the forum maybe 20 different people might give you 50 different opinions
on how it has to be done.

...Starting with '68k had at least VME, but there is no "standard backplane" for a 65816 computer'.
Speaking of it: I remember to have seen an article many years ago in an electronics magazine.
It was about a SBC with VME bus connector, the SBC had two CPU sockets: it was possible to run that SBC
either with a 65816 or with a 68008.

Edit: MC 3/1986 page 48..56.
Eurocard sized PCB, two layers, ca. 40 chips, 4MHz 65816 or 8MHz 68008, 64kB RAM, 64kB ROM, 1*6522, 1*6551.

[BigDumbDinosaur]

Tor wrote:

cbmeeks wrote:

That's right. I forgot about those. There's also the SBC3 too, right? Do you have a link to BDD's version?

He hasn't yet put it up on his website I believe (the POC V1 is there though), so it'll have to be the following thread: viewtopic.php?f=4&t=1688

Yep! I'm woefully behind on documenting my activities. The pages for POC V1 aren't finished either. If only I didn't have to to work...

[cbmeeks]

BigDumbDinosaur wrote:

If only I didn't have to to work...

Work, smerk. How can you put food, water and housing above our '02/'816 needs!

LMAO

[BigDumbDinosaur]

cbmeeks wrote:

BigDumbDinosaur wrote:

If only I didn't have to to work...

Work, smerk. How can you put food, water and housing above our '02/'816 needs!

LMAO

I tried to reduce my food and water intake to an absolute minimum, and looked at large cardboard boxes for housing to eliminate the need to pay property taxes and such. Trouble is, I can't get electrical service to a cardboard box! It just hasn't worked out.

[GARTHWILSON]

ttlworks wrote:

cbmeeks wrote:

Also, the IIgs was purposely kept slow by Jobs (as reported) because he didn't want it to compete with the Macintosh.

Now that's interesting: Commodore didn't want the C65 (3.54MHz 65CE02 core) to compete with the Amiga.

I have a couple of 10MHz 65CE02's that I got around 1990.

Quote:

Hmm... there is a joke, that in hardware design only 5% of the decisions really are technology related.
The rest is tied to politics, logistics, existing tool chains... and dogma.

I would add marketing and the (lack of) ability to foresee the possibilities, good and bad.

[GARTHWILSON]

cbmeeks wrote:

I guess when the 68020 - 68060 came out, the '816 just couldn't compete. Especially when paired with FPU's. Of course, I image an '816 could utilize an FPU too.

True—but what was the price of those? It was getting into the thousand-dollar-processor era.

Quote:

This really makes me wonder what kind of computer could be designed around a 14MHz '816 these days.

I'm sure if the market were there for an '816 PC, they'd make it go a lot faster than 14MHz. 40MHz? 60? Who knows. (The limiting factor would be the outboard supporting parts.) However, although the '816 has its limits, I'm always interested in doing things more efficiently, rather than wasting resources to get the next software product to market, just because "memory/GHz/whatever is cheap." I remember many years ago (late 1980's?) when a new technology was introduced that improved memory prices and speed a lot in one step, and immediately Microsoft was saying, "This is great because now we don't have to be as careful and we can get new software out faster," and what happened is that the user never got the benefit. Boot-up times made no net improvement, the "disc full" messages came up just as often, and there were just as many bugs.

We'll give the video part of the job to Brad. (He'll have to get his parts count down to make the '816 PC affordable though. )