Amiga: the fastest 68k Mac

eharmon

Well-known member
I haven't seen a lot of Mac-on-Amiga topics on here, which is a shame. I've been restoring a 500 and had to give it the ShapeShifter treatment. If you're not familiar, ShapeShifter is the predecessor of both SheepShaver and Basilisk II. Unlike the latter, it's a hardware emulator but not a CPU emulator -- the Amiga also runs on 68k and ShapeShifter can use the native CPU without much loss.

Theoretically, ShapeShifter is the fastest 68k Mac on earth if you've got an Amiga with an '060 accelerator. Sadly that'll cost you 2-3 Macs worth of money. If anyone has one I'd be fascinated to know how quick it is in benchmarking!

Since I didn't want to spend >$1000 on a 1200 and an '060, I took the poor man's option: emulating the only component ShapeShifter isn't, by sticking a PiStorm into my Amiga.

The PiStorm has cropped up here a few times, it basically functions as a bus transceiver for a Raspberry Pi running a 68k emulator. There's always been the option of running the CPU emulator for PiStorm on Linux, but in the past couple years a bare metal emulator has appeared, emu68. This emulator runs without an operating system on top, more directly translating 68k to ARM instructions.

So, more or less, at this point I'm emulating everything on the Mac but the '040 bus. Is that a real 68k anymore? No, not really. But it is fun! And it's fast, other than graphics performance:

IMG_3769.jpeg
Beats a PowerPC!

Impressively fun, that's with the disk image for Speedometer mounted from the Amiga side over SMB over wifi (disk benchmark is against a local disk though).

If anyone has an '060, give it a try, and we can see what it's like without cheating.
 

joshc

Well-known member
If you are going the emulation route, then any modern Mac or PC is going to be a faster 68k than an emulated 68060 ;)

I think for those that are interested, several YouTube videos have covered Shapeshifter.

Beats a PowerPC!
Well yeah, but this is emulation. That Math score does not seem realistic at all.

I've never really understood the interest in Amigas, and I think a fairly accurate analogy is that if Apple had followed the same route as Commodore, then we never would've had Macs that progressed much beyond the Plus.

I'm very sorry, but I had to do this...just because it came into my head and I had to get it out of my head.

1734118853233.png
 

stepleton

Well-known member
Harsh! I found the post interesting, and I've never really used an Amiga or been especially interested in them. Because I am the way that I am, the various games and demos never captured my attention, and instead it's things like the fact that AmigaDOS was originally based on TRIPOS and the presence of the Rexx programming language that grab me instead.

I wonder how complete the emulation is. I was excited about Darkmatter for my NeXT machines, but I was hoping to be able to run Marathon, and the video system isn't compatible. Does the game run on this emulator?
 

Forrest

Well-known member
There’s also PiMega 4, which runs on PC’s, Intel Macs and Raspberry Pi 400. ShapeShifter is discussed at 14:20 at
 

Byrd

Well-known member
Haven’t tried Shapeshifter on my only Amiga - a 1200 with PiStorm lite. Don’t mind the PiStorm one bit - just running EMU68 not full fat Caffeine OS and it’s simply treated as a nice accelerator that works just as well as the $1000 ones.
 

Snial

Well-known member
I've never really understood the interest in Amigas, and I think a fairly accurate analogy is that if Apple had followed the same route as Commodore, then we never would've had Macs that progressed much beyond the Plus.

Occasionally we get Amiga fans on the board so we go through the Mac vs Amiga cycle. This is more common than the Mac vs Atari ST cycle curiously. Commodore's corporate dysfunction was far worse than Apple's, and I think this explains why they struggled to see the Amiga as much more than a balance-sheet filler; whereas Apple saw the Mac as the future of world computing.

The Amiga went down the failed path of bit-planes, which had become obsolete by the late 80s. Note: the early Macs also supported bit-plane type colour, but it wasn't really used and was more easily replaced. But Commodore could have done the same thing: upgrade the Amiga every 24 months (or every 12 months with tick-tock upgrade cycles) and closed the loophole where programmers wrote directly to hardware instead of an abstract API. The PC world managed that, though it took them 1.5 decades, but with the right will Commodore could have done it by the mid-late 80s.

With the right leadership, the Amiga could have become a successful BeBox under BeOS several years earlier. Support the original chipset with hardware access, but offer compelling graphics, sound, multimedia and faster '030, CPUs as they appeared: but only under an API, with the hardware protected via the MMU. For example, use the Ensoniq DOC 5503 would have been a good contender along with 8bpp (palette) and 12bpp (direct) colour modes. Adding textured/gradient lines in the blitter would have made DOOM type games playable on an '030. None of that would have been beyond late 80s hardware.

Apple, in retrospect were almost as bad as Commodore in terms of corporate discipline. The Mac Plus took 2 years to appear and the PCW review was fairly scathing. The Mac II graced the front cover of PCW April 1987, but was already late and should have had an '030 from the start. I remember, the issue contained a poster of the artist's impression of a ray-traced Mac II screenshot, which I had proudly on my wall at Uni. Apple had had multiple, competing Mac projects just to get there. It took them ages to create a portable Mac and essentially messed up the first version until the PowerBooks came along in 1991. The Mac Classic should have appeared in late 1987 and replaced the Plus, but the Plus should have appeared in 1985.

1734173012666.png

Hmmm, Amiga 2000 review in the same issue.

It's difficult IMHO to separate corporate competence from engineering anarchy and market forces as they're all interrelated. Besides the late 80s were a maelstrom of change: 32-bit computers replacing 16-bits; cheap PCs; RISC and the demise of the 8-bit home computer. Not easy to navigate at all. Acorn only survived because of their education market (except they managed to switch completely from 6502 to ARM pretty instantly thanks to EMU65! while Apple struggled to transition from 8-bit to colour 32-bit Macs).
 

NJRoadfan

Well-known member
There was a bit of a cottage industry of emulating the Mac on both the Amiga and Atari ST. Compared to the Mac, they were pretty cheap to buy, plus they had higher resolution graphics vs. an actual Mac Plus! The Amiga could do 640x400 interlaced black and white graphics, while the ST with a monochrome monitor could do 640x400 progressive.

Later on, these emulators could do limited color graphics on the Amiga. While it wouldn't compete with an actual Mac until AGA machines and RTG cards became available, it was still a fraction of the cost of buying an actual Mac II class machine.
 

Unknown_K

Well-known member
If you purchased a Mac II you probably did so because of some kind of expensive Nubus card you needed (DSP, Sound, Video capture).

Kind of hard not to be able to emulate a machine with the same CPU if you can source the ROMs.
 

Daniël

Well-known member
Harsh! I found the post interesting, and I've never really used an Amiga or been especially interested in them.

Agreed, just because the Amiga community can be... difficult, doesn't mean we need to immediately turn difficult the second one is mentioned :p
 

Unknown_K

Well-known member
If Apple had died at the end of the 68K era the Apple community would have been the same as the Amiga one.
 

CC_333

Well-known member
I honestly don't get why the Amiga crowd gets so smug and defensive (or, as
@Daniël so aptly put it, difficult).

What makes the Amiga so much better? Given relatively equal amounts of incompetence and mismanagement at both Commodore and Apple, the Amiga died out and the Mac didn't. Doesn't that say something right there about which was better?

At the very least, Apple did a better job of marketing Macs, and the infamously messy situation that was the Performa series, for what it's worth, did succeed insomuch as they were affordable Macs that came decently configured for what they were and allowed many people to own Macs who otherwise couldn't have afforded them.

Apple's extensive educational outreach with the LC series also enabled schools to provide labs of Macs for their students who may have otherwise not had the opportunity to use or own Macs.

Did Commodore do any of that for the Amiga?

c
 

Daniël

Well-known member
@CC_333 I generally attribute it to the platform's early death.
A lot of the Amiga fanatics never truly got over the fact that Commodore's mismanagement (Irving Gould using Commodore as his personal savings account, and the revolving door of incompetent suits) and eventually demise took the platform and all the dreams of what it could have been with it.

Couple that with several buyouts of the Amiga IP in the 90s by companies that either mismanaged it too, or just intended to sit on it indefinitely, and then the never ending legal battles between companies with bits of IP (Cloanto vs. Hyperion) and attempts at "new" Amigas (and clashes between those insisting on 68ks, FPGAs, or PowerPCs) that will never truly bring the platform back as a competitor to the PC and Mac, and it just creates a poisonous base.

Not all Amiga folk are like this, but it cannot be denied that these cliques fighting amongst themselves are a large problem in their community.
The Macintosh community would have totally been the same, had Spindler managed to kill Apple before Amelio and then Jobs could get the sinking ship back in the right direction (though forum dramas in recent years have definitely brought on some of it).
 

Compgeke

Well-known member
Honestly both Apple and Commodore should've conceeded to HP with their 68k line of workstations, and just ported their OS there. HP was killing it! Like the HP 9000/320 came out in 1985 and already had a 68020. And 1280x1024 8-bit graphics. And ethernet. And 1 MB RAM. Apple and Commodore could've spent their development money subsiding the HP hardware which was technically superior. Plus you could switch from Unix at work to BASIC at home to Mac for the kid's homework then Amiga for the video editing.

Commodore later wanted to move to PA-RISC, so maybe they learned their mistake that they should've been licensing the HP 9000 instead of dealing with lowly consumer hardware.
 

Mk.558

Well-known member
Features are nice to have, but cost is a factor too. In the late 80's there was a big memory price spike and offering a computer with 8MB of memory was fabulously expensive.


I like the SE/30, many other people love it too. But let's be honest: the base spec was around 5000$ US, and that was for 1MB of memory and a 40MB hard drive. That is a lot of money, in late 1980s money values, and there was a recession at the same time too in Britain, Japan and the US. It cost about as a Macintosh II. The IIci, another machine I rate fairly highly, cost even more!
 

Forrest

Well-known member
I started on an Atari ST in 1985. $499 for a 520ST, 400K 3.5 inch floppy and a 640 x 400 monochrome monitor. The Amiga may have been great, but out of my price range at the time. Upgraded the ST over the years and enjoyed running Mac software with a Spectre GCR running System 6.0.5. It was an easy switch to a Mac when Atari went under.
 

tokenalt

Member
Honestly both Apple and Commodore should've conceeded to HP with their 68k line of workstations, and just ported their OS there. HP was killing it! Like the HP 9000/320 came out in 1985 and already had a 68020. And 1280x1024 8-bit graphics. And ethernet. And 1 MB RAM. Apple and Commodore could've spent their development money subsiding the HP hardware which was technically superior. Plus you could switch from Unix at work to BASIC at home to Mac for the kid's homework then Amiga for the video editing.

Commodore later wanted to move to PA-RISC, so maybe they learned their mistake that they should've been licensing the HP 9000 instead of dealing with lowly consumer hardware.
It wasn't superior tech it was higher prices, the HP 320 had a starting price of $8510. The only thing that stopped Apple or Commodore from using those high end parts was the fact that no home user would pay those prices.
 

Unknown_K

Well-known member
Amiga was cheaper than a Mac, but the ST was cheaper than the Amiga. I have a 1040ST and both Atari monitors (color and mono) and the mono would have worked great as a cheap Mac.
 

Arbee

Well-known member
Commodore bet everything on custom hardware, which is the classic "how it started/how it's going" meme. Off-the-shelf hardware kept getting better and cheaper while custom chips kept getting more expensive. This is the same thing that killed arcade games and made game consoles go to standard components.

By contrast, the Mac and PC bet everything on software from day 1, which enabled a lot of flexibility. Well-behaved System 1.0 programs can run on 9.2.2 and even OS X Classic mode. Well-behaved Windows 1.0 programs can run on Windows 10. On the Amiga, most games were written for the Amiga 500, and couldn't even boot on subsequent models.
 

Snial

Well-known member
Lots of interesting comments exploring various theories. To me it looks like there's elements of truth in all of them. Particularly:

<snip> Commodore's mismanagement (Irving Gould <snip> dreams <snip> buyouts of the Amiga IP <snip> "new" Amigas <snip> creates a poisonous base.
As a veteran Mac fanatic, totally true that our kind of zealotry is very off-putting (I've been very guilty of that!).

<snip> Commodore later wanted to move to PA-RISC <snip>
This is a revelation! I have a PA-RISC HP 715 (or 725/735 or that kind of thing)!

I started on an Atari ST in 1985 <snip> Amiga <snip>out of my price range <snip> enjoyed running Mac software with a Spectre GCR running System 6.0.5. It was an easy switch to a Mac when Atari went under.
ST as a springboard to Mac, cool!

Amiga was cheaper than a Mac, but the ST was cheaper than the Amiga <snip>
Maybe the Amiga / ST market split harmed the long-term chances of both platforms.

Commodore <snip> Off-the-shelf hardware kept getting better and cheaper while custom chips kept getting more expensive. <snip> By contrast, the Mac and PC bet everything on software from day 1,<snip>
I don't really buy into that theory, partly, because I think that the history of PCs benefitted from plenty of custom chips/boards/FPGAs etc. But it may be true and particularly insightful. However, I agree that software compatibility across models is a real help and as I see it, the Amiga largely failed in that respect (and the Atari ST struggled too). In some ways it's a case of worse-is-better: original Mac hardware was, frankly, rudimentary. It was classic Apple ][ "minimise the hardware and make the software sweat" principle. It works, because a low base is an easy common denominator for future support.
 

Arbee

Well-known member
I don't really buy into that theory, partly, because I think that the history of PCs benefitted from plenty of custom chips/boards/FPGAs etc. But it may be true and particularly insightful. However, I agree that software compatibility across models is a real help and as I see it, the Amiga largely failed in that respect (and the Atari ST struggled too). In some ways it's a case of worse-is-better: original Mac hardware was, frankly, rudimentary. It was classic Apple ][ "minimise the hardware and make the software sweat" principle. It works, because a low base is an easy common denominator for future support.

The difference is that PCs and Macs benefitted from a lot of third-party ASICs and FPGAs, but the platforms were not defined by those specific chips or compatible successors the way the Amiga was. And Apple/Dell/HP/Compaq weren't paying the full development costs for them the way Commodore was. Most 68k and PPC Macs had custom chips, but they were largely gate arrays rather than full custom silicon and generally had different combinations of the same set of building blocks.

The other element is that the custom chips themselves are becoming more software-driven. The current AI craze is predicated almost entirely on graphics chips in the early 2000s deciding to become massively parallel CPU clusters to emulate the pipeline defined by Pixar's Renderman software. Networking hardware with significant processing capability of its own to offload from the main system is becoming a big deal in the server space now too.
 
Top