ARM or x86 with FPGA emulator

[ElPolloDiabl]

This might be good as an AROS bounty. Otherwise just a hypothetical.

Has there even been an attempt to use an x86 or ARM as a co-processor for the Amiga?

[ElPolloDiabl]

P.S. Someone start a kickstarter to have OS4.1 ported to the much cheaper ARM platform.

[Hattig]

Hyperion just isn't interested in porting AmigaOS onto a cheap, mass market hardware solution.

AROS on ARM is the best option, with a 68k emulator for native app compatibility (hopefully still calling the native AROS libs, like OS4's Petunia emulator).

An FPGA for native hardware compatibility is a nice thought, but would turn the hardware from cheap commodity into bespoke expensive.  So a perfect match for this community.

[takemehomegrandma]

AROS on ARM is the best option, with a 68k emulator for native app compatibility (hopefully still calling the native AROS libs, like OS4's Petunia emulator).

Petunia (OS4) or Trance (in MorphOS)  is more of a binary translator/recompiler than it is a emulator/simulator; it basically recompile the 68k bytestream into native PPC code. This makes sense on a big endian HW architecture like PPC, because it makes it possible to mix "68k code" with PPC code in the same environment and run old and new programs (and OS components as well, like ARexx for examples) in the same environment as they were no different to each other. In fact they *aren't* different, both 68k and PPC apps are all de-facto PPC native at runtime, being scheduled by the same scheduler, sharing the same memory and system resources, etc. In this way, both MorphOS and OS4 *is* the native Amiga environment, where both new and old applications runs side by side, but on a different HW platform.

This isn't really possible on a little endian platform though, and AROS has a different approach where UAE is used to emulate a complete Amiga machine (with 68k CPU, custom chips and the whole shebang). The upside of this is an improved compatibility with very old SW and/or HW hitting applications, but the downside (compared to the MorphOS/OS4 approach) is that the Amiga apps runs "sandboxed" inside this emulated Amiga, and the AROS native ARM/x86 apps runs outside this box under AROS.

An FPGA for native hardware compatibility is a nice thought, but would turn the hardware from cheap commodity into bespoke expensive.  So a perfect match for this community.

A terrible idea if you ask me. There are mainly two groups of Amiga users today, those who are here for the "classic", who are more of retro fans today, and those who are here for "NG", which is more about evolution than retro. The retro people use real Amigas, or emulators, or "new classic" HW like Minimig.

A product like you are talking about would not suit the retro fans, since they are more interested in "the real deal" and 100% compatibility (and like it or not, new HW compromises backwards compatibility). And the "NG" fans will feel the platform is being held back by 20 year old museum level technology. So it would be stuck in the middle, not really satisfying any of the groups.

IMHO of course!

[OlafS3]

I do not feel "retro" like some of the "NG" fans always insist. For me "68k" is only a compiler target and running in emulation beats many of the so called "NG" options. Aros Vision f.e. inherits the same limitations that all AROS versions have. It is part of a bigger family :-). So for me "68k" is not retro and has the advantage to run almost anywhere.

What are the advantages of NG except faster 3D (if drivers would exist)?

[ElPolloDiabl]

I suppose I was talking about the PowerUP cards, but that usually ends up being a fast computer with a slow computer tacked on.

[bloodline]

I suppose I was talking about the PowerUP cards, but that usually ends up being a fast computer with a slow computer tacked on.

This is the problem.

One could take a nice fast ARM chip, as used in a tablet or smart phone. These things are usually multiple core chips with a gig or so of RAM, and come in at a power budget of of only a couple of watts and don't produce much heat.

This makes them ideal for use in an Amiga accelerator... But they also tend to have a graphics core included (much better than AGA), also they have high def audio outputs, and they have USB as standard now... So why bother making all that effort to mate it up with an Amiga motherboard, the only parts you would probably use are the keyboard and disk drive...

[Linde]

Makes me think of http://www.amigahistory.co.uk/inside.html

[IanP]

An ARM CPU card for the FPGA Arcade would make an interesting addition to it. It would give you a lot of options. You could configure the system to work as a 68k AGA+ machine with the ARM emulating the 680x0 (assuming it would outperform a softcore) or the ARM could be used as a coprocessor to offload jobs like video, audio, jpeg decoding and provide networking and USB features or you could configure it to work as an AROS on ARM NG Amiga.

If Hyperion had any interest in porting OS4 to anything else they'd have done it by now.

Because MorphOS has also remained shackled to PPC it's not something I've paid much attention to so I don't know how open they would be to porting to another architecture.

[psxphill]

This makes them ideal for use in an Amiga accelerator... But they also tend to have a graphics core included (much better than AGA), also they have high def audio outputs, and they have USB as standard now... So why bother making all that effort to mate it up with an Amiga motherboard, the only parts you would probably use are the keyboard and disk drive...

I sort of agree, but the phase 5 boards had a graphics card option too.
 
I quite like the idea of being able to run a 68k & PPC emulator on a fast ARM processor and being able to use either AGA or the accelerators graphics output.
 
Pure cpu emulators designed to run full speed should be able to run quicker than one designed to emulate AGA etc as well.
 
I guess you'd boot from AROS for ARM and have it so that could select between booting an ARM workbench, which could run 68k apps in AROS using headless mode (like juae) or startup a 68k task that just booted from a 68k kickstart. Quite what the hard drive of a dual boot system would look like is another matter, ideally if you disable the accelerator it would use the onboard 68000 (A500/A600/A2000), 020 (A1200) or 030 (A3000/A4000CR) and it would still be able to use the onboard 68k kickstart and the 68k workbench from the hard drive. A4000 is the odd one out, there is no onboard CPU.

[matthey]

An ARM CPU card for the FPGA Arcade would make an interesting addition to it. It would give you a lot of options. You could configure the system to work as a 68k AGA+ machine with the ARM emulating the 680x0 (assuming it would outperform a softcore) or the ARM could be used as a coprocessor to offload jobs like video, audio, jpeg decoding and provide networking and USB features or you could configure it to work as an AROS on ARM NG Amiga.

The FPGA Arcade has an ARM processor. It is low end so probably wouldn't help with emulation much but it can be used to offload some tasks much as the version of the MiniMig with ARM.

[ppcamiga1]

Adding x86 or arm to the fpga to emulate 68k/powerpc is a bad idea.

No one will make software for the emulator.

Amiga 68k is not a bad computer, but 68k is too slow.
 
What we really need is to add powerpc to fpga.

Fast 68k does not exist, the mythical 68k from natami exists only in a sick imagination of gunnar von boehn.

x86 and arm as litte endian processors do not allow for integration with existing software.

SPARC, Itanium, Z * from IBM are too expensive.

MIPS, SH4 are too slow.

PowerPC is the only reasonable solution for the Amiga in an FPGA.

[matthey]

Adding x86 or arm to the fpga to emulate 68k/powerpc is a bad idea.

No one will make software for the emulator.

Amiga 68k is not a bad computer, but 68k is too slow.
 
What we really need is to add powerpc to fpga.

Fast 68k does not exist, the mythical 68k from natami exists only in a sick imagination of gunnar von boehn.

In other news Gunnar, Jens and Chris just tested the new Phoenix Demo on Majsta's Vampire 600 using ECS chipset, 16 bit fast memory and a too small Cyclone II fpga:

http://www.apollo-core.com/bringup/roto64.jpg

Not only is the demo working but that 31 in the upper left hand corner is the frames per second (fps). Here are some other results:

 A4000/040 = ~7 fps
 A4000/030 = ~3 fps
 A600/TG68 = ~3 fps
 A600/TG68-020+cache = ~5 fps
 AMIGA 1000 (fastmem) = <1 fps
 A1200 68030@50 = ~5 fps
 A1200/1260@80MHZ = ~19 fps
 3000T CSMK3 68060@75MHz = ~20 fps average (15-28fps)

A600/Phoenix = 31 fps

The CSMK3 68060@75MHz is my Amiga. The Phoenix demo is here:

http://www.apollo-core.com/phoenix_demo4

The reason why the 68k is slower than other processors is that no large companies with deep pockets are developing it. Of course the 68060 is slow, now .

PowerPC is the only reasonable solution for the Amiga in an FPGA.

There is a comparison of the old Apollo core with the PowerPC 440 core in fpga:

http://www.apollo-core.com/index.htm?page=performance

I believe the 68k register memory architecture is better able to take advantage of the large memory bandwidth of an fpga than the load/store architecture of the PowerPC (and most other RISC fpga cores). High clock speeds are not necessary for strong performance when the processor can work directly in memory where there is plenty of memory bandwidth. This is even better when it can work in memory in parallel. We will have to wait for the Apollo core for this as the Phoenix core is just a single integer unit (68060 has 2 integer units).

[ElPolloDiabl]

That speed... is it because of the better memory speed or is it because the core has been optimised?

It's impressive.

[dammy]

Adding x86 or arm to the fpga to emulate 68k/powerpc is a bad idea.

No one will make software for the emulator.

Amiga 68k is not a bad computer, but 68k is too slow.
 
What we really need is to add powerpc to fpga.

Fast 68k does not exist, the mythical 68k from natami exists only in a sick imagination of gunnar von boehn.

x86 and arm as litte endian processors do not allow for integration with existing software.

SPARC, Itanium, Z * from IBM are too expensive.

MIPS, SH4 are too slow.

PowerPC is the only reasonable solution for the Amiga in an FPGA.

Just run 68K code in UAE and use AROS for the ARM/x86.   Might not be the perfect solution, but considering cost and man power, it's the only viable solution.