Opinion: PPC, 68080 and other

[pix]

Not wanting to start a flame war with my first post on this board, but I'm hoping to get some solid opinion on the best Amiga hardware direction to take.

For where I am coming from, after experimenting with various Amiga offerings over the years, emulation options are a given. Love digital preservation and it does the job + more, but its still emulation.

Pimiga is a great combo of emulator and hardware and on a Pi400 it almost captures the feel of having an Amiga very well, so I'm happy with that end of things.

If I was going to opt to use an OS on off the shelf hardware, while I liked the look of MorphOS and tried it on an iBook the idea of owning Apple hardware to run Amiga makes me feel a bit.... dirty.

I was enthused about AROS. If I must use off the shelf, then x86 is as good as any and AROS does feel very Amiga. Unfortunately it was going well but now seems (to me) to have stalled for some time.

Of course I realize you can take an original Amiga and transform it with accelerators and I'm glad some people do, but I don't think its for me.

Which means if I am picking something that runs on consumer grade hardware then I am really drawn to OS + alternative hardware. Coming across the 68080 FPGA rekindled my interest, but as I read about the performance I was a bit disappointed and it doesn't run OS4.1 correct? So I have to choose between OS4 and 3. I love the look of OS4 and I'm not against PPC hardware as a way forward, but its a shame the Amiga is so sharded by CPU and OS and so on.

I'm not interested in ARM processors at all really.

I also don't really understand why the PPC and 68080 are bounded by clock speed. Why can't the PPC push to even the triple core 3GHz of an Xbox 360? Or any is the 68080 clocked at 80-odd MHz? Why not 500? I don't get it.

Would really hope the Amiga will one day get standardized on OS 4 with either Moto or PPC redesigns at higher power and its own chipsets to really offer Amiga developers a way to flex the differences of an Amiga against anything else.

In a world where we have to choose between boring ol x86 and an endless stream of ARM, we need that!

[Samurai_Crow]

Re:GHz 68080
The FPGA used in the Apollo standalone models is bounded in size and clock speed due to being a low-end prototyping chip. Once the bugs are worked out it could theoretically be mass produced as a fixed function ASIC SoC in GHz clock speeds but still would not be competitive. The 68080 is a 4-stage dual-issue CPU only a generation past the 68060, equivalent to a Pentium 2.

Re:multicore Amiga
The Amiga OS supports only single-threaded operation, the exception being that the chipset runs independently. Modern Amiga emulators are limited to 2 threads maximum because of that fact. Amiga itself was advanced back in the day because of its parallel chipset design but now it's hopelessly behind the times.

[grelbfarlk]

There is no reason the, a Amiga can't have a Cell PPC or some 2GHz+ multi-core PowerPC.

Nobody has done it because it costs money and time, and licenses.

If you want to run AmigaOS 4 on a PPC you have to get a license from Hyperion or A-Eon(if you want to use their drivers).

If you want to run AmigaOS3 with your 68k and add a PPC to it running WarpOS, you can do that, it's pretty well documented. But most of the OS is running on the 68k. This maybe matters or or not, it depends on your sincere religious beliefs or the existence of the AGA or ECS chipset and 68k being there.

[Thorham]

Originally Posted by Samurai_Crow

The Amiga OS supports only single-threaded operation

Don't you mean single core?

[Samurai_Crow]

Originally Posted by Thorham

Don't you mean single core?

The only instance of hyperthreading that has been supported on Amiga was the software implementation of the blitter on the Vampire v2 that used AMMX vector ops in a nested loop. As such it was an exception to the single thread rule.

I've heard of some rumors of AmigaOS 4.x being able to use "slave" cores for some concurrency on multicore PPC processors but I think that's in beta testing, if even that.

If you're referring to tasks and processes in the multitasking kernel, they are not allowed to be concurrent, due to high task priorities being required to block lower task priorities.

[Gorf]

Originally Posted by Samurai_Crow

The Amiga OS supports only single-threaded operation, ...

As always when it comes to SMP-support I have to mention that it has successfully be done for AROS.

The SMP build for pc-x86_64 works ... not very stable, but it works.

https://amitopia.com/aros-open-sourc...pport-to-work/

[Pyromania]

68080 all the way! Vampire V4 is a nice upgrade path to classic Amigas. Amiga is a retro market now so does not need the latest and greatest CPU. SAGA chipset with 68080 is a nice bump up in CPU and custom chipset. Also the price is reasonable compared to PPC which is a very dead tech in 2024.

[Samurai_Crow]

Originally Posted by Pyromania

68080 all the way! Vampire V4 is a nice upgrade path to classic Amigas. Amiga is a retro market now so does not need the latest and greatest CPU. SAGA chipset with 68080 is a nice bump up in CPU and custom chipset. Also the price is reasonable compared to PPC which is a very dead tech in 2024.

Not only is a fast CPU not needed for Amiga software, but PC ports from open-source projects are the only programs that take advantage of modern hardware.

My G4 Mac Mini runs MorphOS smooth as butter but has to rely on EUAE to run Amiga software that needs chipset functions. Otherwise I'd recommend it as being a cheap aftermarket system for running Amiga stuff, now that it has graphics acceleration with shader support from beta drivers.

AmigaOS 4.x has proven to be a dead-end for its promise not to break backward compatibility ever. SMP won't make an appearance there. Only "green" threads are possible on it. (Green threads are where slave cores run a task to completion and then free up the core.)

@Pyromania
How cheap is an Apollo core system? I got my G4 Mac Mini used for $50 but it was missing a power supply. I robbed the power supply off my Intel Mac Mini dual-core and registered MorphOS for whatever it cost to do so. The system was up and running for less than $200.

If the Apollo computer can't match that price point, maybe Apollo isn't for the faint of heart. Last I checked the latest Apollo core system with 512megs was as much as a new computer anywhere. Now they have a 1 gig version? It must cost a pretty penny to buy a new "Amiga". (€500 here in the USA is about what I paid for my A1200 back in 1993. That's still less than the latest Apollo standalone.)

[TCD]

Originally Posted by Samurai_Crow

How cheap is an Apollo core system?

Without taxes the core system boards are around 550$.

[Samurai_Crow]

Originally Posted by TCD

Without taxes the core system boards are around 550$.

Thanks, that's what I thought.

[Karlos]

Not only is a fast CPU not needed for Amiga software, but PC ports from open-source projects are the only programs that take advantage of modern hardware.

Well now, tell that to 3D rendering software. I loved messing around with those back in the day, but my 040 could still take a couple of days to render a scene properly (if I wanted proper ray traced lighting and a decent resolution anyway). I used to use OctaMED sound studio a lot too. Try running multiple channel mix mode with smoothing enabled. You soon crush an 040 and an 060 can't go that much further.

Plus, the PiStorm is literally the only way to play AB3D2 at fullscreen 1x1 at framerates people think is necessary today. A lot of effort has gone in to just make it playable on a 68060.

[pix]

Originally Posted by Samurai_Crow

Re:GHz 68080
The FPGA used in the Apollo standalone models is bounded in size and clock speed due to being a low-end prototyping chip. Once the bugs are worked out it could theoretically be mass produced as a fixed function ASIC SoC in GHz clock speeds but still would not be competitive.

The 68080 is a 4-stage dual-issue CPU only a generation past the 68060, equivalent to a Pentium 2.

Re:multicore Amiga
The Amiga OS supports only single-threaded operation, the exception being that the chipset runs independently. Modern Amiga emulators are limited to 2 threads maximum because of that fact. Amiga itself was advanced back in the day because of its parallel chipset design but now it's hopelessly behind the times.

Interesting. Personally I just think it would be great to take sac the 68080 base and Amiga chipsets as far as possible. For me it doesn't have to be "competitive" in the ultimate sense, but with appropriate instructions, accelerators and the fastest possible single core performance it could become a comfortable daily. If it has the grunt to push HTML5 around in a decent browser and h265 video acceleration that's half the battle. All the better if its an Amigan implementation. Basically what an emulator can do - just get it pushing out as fast as possible

[pix]

Originally Posted by Pyromania

68080 all the way! Vampire V4 is a nice upgrade path to classic Amigas. Amiga is a retro market now so does not need the latest and greatest CPU. SAGA chipset with 68080 is a nice bump up in CPU and custom chipset. Also the price is reasonable compared to PPC which is a very dead tech in 2024.

I'm inclined to agree with this. I'd see the PPC option as more interesting if it was advancing somewhat, otherwise the Motorola extensions and building up what an Amiga ultimately could have been when its made as fast as possible in hardware seems more interesting

[Bruce Abbott]

Originally Posted by pix

Would really hope the Amiga will one day get standardized on OS 4 with either Moto or PPC redesigns at higher power and its own chipsets to really offer Amiga developers a way to flex the differences of an Amiga against anything else.

What would a 'higher power' OS 4 'Amiga' offer developers that other platforms don't?

In a world where we have to choose between boring ol x86 and an endless stream of ARM, we need that!

I agree that x86 and ARM are boring. But OS 4/PPC isn't any better IMO - still the same exploitive business model that keeps you on the treadmill, running faster and faster just to keep up. That's why I choose to continue using my A1200 with 50MHz 030. I also have an A600 with Vampire, but I find the higher power systems less interesting. Finally I have a stock A500 with 1084 monitor to preserve that true retro experience.

For me, nothing beats real 1990's Amiga hardware doing the stuff we were doing (or wanted to do) back then. So many games, so many programming challenges, so many hardware projects I want to do - I doubt that I will live long enough to complete a fraction of them!

If you chase the high end you will never be satisfied. No matter how high powered an OS 4 'Amiga' is it still won't match a PC. The OS and apps will always be a little clunky and not do the job quite as well, and you won't be able to do half as much despite spending megabucks on it. So why not just use a PC for that stuff?

[xavlahaye]

Originally Posted by Bruce Abbott

That's why I choose to continue using my A1200 with 50MHz 030.

Same here I tend to think that the 1230 50 Mhz is the smoothest experience.

Originally Posted by Bruce Abbott

If you chase the high end you will never be satisfied.

Exactly, that's what I did for a while. There are always things that won't work and apps for high-end amigas where/are developed with fewer ressources and tend to feel clunky/buggy/unfinished...

[OlafSch]

Originally Posted by pix

I'm inclined to agree with this. I'd see the PPC option as more interesting if it was advancing somewhat, otherwise the Motorola extensions and building up what an Amiga ultimately could have been when its made as fast as possible in hardware seems more interesting

I see it similar as Bruce. Even if you have a somewhat modernized amigaos it still misses resources, expecially modern software and drivers. That cannot be fixed easily. That shows aros very transparent. That will be different with axruntime and the merge with linux but all standalone solutions currently have similar problems.

[TCD]

Originally Posted by xavlahaye

There are always things that won't work and apps for high-end amigas where/are developed with fewer ressources and tend to feel clunky/buggy/unfinished...

There might be a reason for that.

Just out of curiosity: The Vampire 'strain' of Amiga add-ons is on pause it seems? Even the website didn't get updated this year.

[eXeler0]

As someone who has a truckload of real Amigas I’ll say that the path forward I’ve chosen for the ”High end” is PiStorm. (I had a Vampare for a while and I also have a Mac Mini with MorphOS)
The performance / dollars of the piStorm is unmatched when compared to other systems where real hardware is somehow kept (meaning, you still keep your actual Amiga from back in the day). And for each new Raspberry Pi generation you get a potentially faster Amiga.
And going back to real hardware takes as long as it takes to pull out the PiStorm card (or replacing it with a 030 or something appropriate).

[aeberbach]

I would love to have a 68060 but rarity and cost make it impossible. I don't much see the point of PowerPC. If you just want classic Amiga the Pistorm seems to be a great option. I have vanilla 3.2 Workbench and have added the Pistorm, it's as good as a 68060 ever was with Emu68. Haven't got RTG configured yet but am looking forward to it. Of course there will be edge case incompatibilities I don't know about yet but it is on a cost level with 68030 and far superior!

[minator]

I've only recently been looking at the Amiga again after a big gap. I got a new powerful Mac to play with Ai and it's so powerful it got me thinking about how well emulation would work on it and that got me looking at the Amiga scene again...

For the future, I think there are a few options but it depends on what you want to do.
For Classic users I think PiStorm seems to be the way to go, cheap, powerful and ever increasingly more powerful.

The Vampire project looks really interesting, I remember it from many years ago and I thought they were mad, I always thought they were being far too ambitious, but hats off to them for actually getting it done. I think it's a really cool idea but it's expensive and an FPGA will limit the performance, it has a limited market.

For the higher end PowerPCs it's more complicated, PowerPC is pretty much a dead end now so the PowerPC machines have effectively become a kind of second round of classics.
To keep up with more modern processors they'll need to switch to x86 or Arm64.
If not, they'll be in the potentially embarrassing position of being overtaken by the Raspberry Pi ...if that hasn't happened already, I imagine the Pi 5 is giving most PPCs a run for their money.

Arm didn't look very interesting for the desktop until Apple did the M series, recently Qualcomm released the X-Elite chips pushing Arm into the PC market in a much more serious way than before. The x86 emulators are far better now and these machines are getting some pretty good numbers. However there are many more on the way and there's a whole who series of Arm chip manufacturers who could appear soon.

However, porting an OS is a big job and is there the will or resources to do it?
What about Aros? That actually looks like the best option as it's already multi-platform so is best positioned for use with all these newer chips.
I think a hosted (i.e. running on top of another OS) is a good option as it won't require hardware drivers. Yes there is a performance cost to doing this, but modern CPUs - even single cores are so fast now, I think you'll be hard pressed to notice any difference.

What's really interesting is this axruntime project which I'd never heard of before. That'll not only give access to modern hardware, but I'm also assuming it'll give access to multiple cores as well. Amiga will effectively become an environment - using the resources of Linux (etc.) to give you a fully modern system - like how Android uses Linux as a base.

So, a modern system could be an axruntime based Aros environment with something like Amiberry for classic / 68K software.
Built for Arm64, that'll run on anything from a Raspberry Pi 3 up to the new X-Elite machines (and presumably Macs with some effort).

Not the ideal situation but it would be available, high performance, and cheap.