Quote:

http://ultimateppc.nl/


Early days, but a possible MorphOS intrest?

My thoughts:

https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=6465&forum=11&start=91

Would love to see MorphOS on classic again.

I wouldn't hold my breath on that at all:

1. This project probably will be never ending story
2. MorphOS Team already declared what do they think about supporting classic hardware.

EDIT: [Wolfmode on] http://www.amiga.org/forums/showpost.php?p=693973&postcount=26 [Wolfmode off]

[ Edited by pampers 23.05.2012 - 11:14 ]

Really cool if it comes out. The 1541-Ultimate is a really nice piece of hardware. Will I buy another Amiga classic? Probably not if no MOS support, but we'll see.

I dont know what other developers think but I doubt it. In my opinion real Amiga (real Amiga as in Amiga 500 and Amiga 4000 models) are best with a 68k CPU and AGA.

Amiga 3000/4000 are so rare today it is not worth of effort... I rather pass and encourage our AROS friends port their OS into it.

The board seems to a rather complete ppc board + a 68k Amiga expansion board. The 68k part should use the peripherals of the P1013.
As far as I understand the ppc board operates rather individually (dunno whether the attached A3/4k will be visible to the ppc board at all). That makes that board a small ppc board. Which generally is interesting for MorphOS. But the e500V2 core is yet not supported by MorphOS as aren't the rest of the SoC compounds and the onboard gfx chip.
To support that board a lot of new drivers would be requireded. Rather doubt that effort would be worth the benefit. I mean (no disrespect, just a reality check) how long does/did it take to support Efika, Mac mini, Powermac or Powerbook? Resources are limited, better spend them on more promising stuff (G5, e6500 based devices or ISA switch).

I don't think this board is high on the MorphOS priority list, but one should keep it on radar and look where thet project actually wil go to.

> The board seems to a rather complete ppc board

Can audio hardware and 3D graphics hardware attached to the A3000/4000 be used from the PPC side of this expansion board? To me this would seem crucial for using the PPC as the main CPU.

> the e500V2 core is yet not supported by MorphOS

Even if MorphOS was ported to the e500v2 core, existing MorphOS binaries that use the FPU wouldn't run. And it seems the e500v2's "embedded FPU" overlaps with AltiVec as well.

http://en.wikipedia.org/wiki/PowerPC_e500#e500v2

> as aren't the rest of the SoC compounds and the onboard gfx chip.

Onboard gfx chip? Is there something else on the board than the P1013's on-chip DIU?

There would need to be a compelling reason for someone to use this board and an original Amiga computer to run MorphOS3.x, instead of a cheaper faster solution, such as a 1GHz, or faster G4 PowerMac, or G4 MacMini, or soon the G4 PowerBook's that will be supported with the next release of MorphOS.

The integration of the onboard Amiga Custom Chipset for example and the way it is implemented would have a lot to do with how successful it would be, or not. From past experience with the Phase5 boards that had both 680x0 and PPC CPU's on them and the slow context switching that caused such a performance penalty, that as good (no, make that great) as those boards and software were and still are, they are no match for what has already replaced them.

I am not opposed to the project and admire the effort, but they would need to be a huge improvement above and beyond what is already available to create new interest and sales of these boards, instead of just using an A4000 for 680x0 software and MorphOS2.7 (or OS4.x if that is your preference) for PPC Amiga software.

I think you all underestimate the "I want that"-factor. I do not know how many they will produce, if ever, but if it's more or less affordable... they will all find new owners. Just as subways, indivisions, denebs and mediators and old AmigaONE en Pegasos boards still get sold...

Anyway I will buy one for sure. If it can replace the PicassoIV and the Deneb and my current accelerator in the A4000 and sell those, I probably do not need so much extra cash anyway. Well I hope that anyway

> From past experience with the Phase5 boards that had both 680x0 and
> PPC CPU's on them and the slow context switching that caused such a
> performance penalty, that as good (no, make that great) as those
> boards and software were and still are, they are no match for what
> has already replaced them.

You are probably aware of that, but just to make sure as your comment reads ambiguously to me:
Context switches are only involved when using the m68k CPU as the main CPU and the PPC as the coprocessor, such as done via WarpOS or PowerUP on OS3. When using the PPC as the main CPU (via MorphOS, OS4, Linux/PPC etc.), the m68k CPU only does the initial start-up of the machine and is then put to sleep (after all, in MorphOS/OS4 it's being emulated by Trance/Petunia), which means there are no context switches whatsoever involved when running a PPC OS on the PPC CPU of a combined PPC/m68k expansion board.

Addendum:

> Even if MorphOS was ported to the e500v2 core, existing MorphOS
> binaries that use the FPU wouldn't run.

Confirmation:
https://morph.zone/modules/newbb_plus/viewtopic.php?forum=11&topic_id=6465&start=92

MorphOS interest? What would be the point?

For me it's a hobby, I don't need a point ... I would buy it.

Andreas, from amiga-news.de:
Quote:

Last item on that list. Reads as if they would put a chip onboard, but probably you are right and the integrated DIU is meant.

Be it this or that, a driver would be required anyway.

[ Editiert durch Zylesea 24.05.2012 - 21:39 ]

Quote:

And I think you overestime the targetmarket, A3000 and A4000. Anyone getting their hands dirty on this must be declared insane :)

> from amiga-news.de:
>
>> [...]
>> Integrierter 2D-Videochip mit DVI-Anschluss

Apparently, this is just a distorting translation of the original "Embedded 2D video with DVI connector" statement, which as you can see says nothing about a separate "chip".

> probably you are right and the integrated DIU is meant.

Yes, I think so.

I think there are more A3/4000 alive as MorphOS licences sold ... so who's insane ? Developing anything amiga seems to be insanity in general ... but still it happens for some reason.

Oh well, let us wait and see...

Quote:

From the project website.. so in otherwords probably never going to happen..

magnetic,
Quote:

I have my reservations this will come out, and almost certainly not before the X-mas 2012. The only thing that stirs my interest is it by the guy who did the 1541-Ultimate which is a real product which has shipped in multiple revisions. The 1541-Ultimate is a pretty nice achievement itself and quite popular in C= circles. This PPC board appears to be a bigger undertaking, so we shall see.

@hooligan,
Quote:

...and of those who still actively use these kind of machines, I think it's reasonable to assume that many *already have* 68k accelerators faster than this one (i.e. 68060 or 68040 > 33MHz), those who needs USB already got it, same with network and HDD/CD. Had it come with a 68060 @ 100MHz it would offer a reason to bother, but as it is now, I think it could be a downgrade for most people. And when the PPC used in this can't run existing Amiga/MorphOS PPC code, it kind of tips it over the edge to redundancy.

I agree with Zylesea, a 8610 would have been a much better start:
* Up to 1333 MHz performance and 3060 MIPS
* DDR/DDR2 SDRAM memory cotroller @ 533 MHz (1066DDR)
* Altivec (possibly the fastest implementation on any "G4" from what I heard)
* Integrated 2D display controller supports up to SXGA 1280 x 1024 resolution and 24 bits per pixel (See it play 1080p streaming video, discussion about it here)
* It has PCI-e, PCI2.2, and many other I/O interfaces that probably could be glued to the Amiga CPU slot

* SATA, USB, high speed network, audio, modern GFX cards, would be easy
to obtain, either through PCI/PCI-e expansion slots, on board controllers, a mixture of them both, or through the use of a southbridge. Think of these HW specs, but in a shape and form that suits A3000/A4000 layouts.

* Skip the 68k CPU entirely, it only adds cost and complexity. Instead, handle it all on the PPC side. Make a 68k->PPC JIT translator, like Trance in MorphOS; its 68k performance is incredible. Store it in some ROM on the board, and let this "CPU" run the 68k show from moment of powering the system up. All this with the PPC chip completely transparent from the 68k side. Classic environments has no need, nor any want, for PPC OS's anyway (and Vice Versa of course).

Then we are talking *serios* accelerator boards, worthy of the name "UltimatePPC"! Of course, making a 68k JIT like Trance takes its programmer, as well as a couple of years perhaps.

And of course, none of this will happen, just thinking out loud...

[ Edited by takemehomegrandma 25.05.2012 - 09:19 ]

> a 8610 would have been a much better start:
>* Up to 1333 MHz performance

1333 MHz MPC8610 parts are no longer manufactured. It goes up to 1067 MHz now.

http://www.freescale.com/webapp/sps/site/prod_summary.jsp?code=MPC8610&tab=Buy_Parametric_Tab

> * Altivec (possibly the fastest implementation on any "G4" from what I heard)

It's the very same implementation as in MPC7448 and MPC8640(D)/MPC8641(D) from what the Freescale docs say.

> * [...] audio [...] would be easy to obtain

Obviously, as it's already integrated in the MPC8610 (I²S/AC'97).

This thread is just conjecture. There are reasons why nobody has brought a CLassic Amiga PPC board to market since Cyberstorm/blizzard series...

its all fantasy talk

Do you have any idea how much fantasytalk came to reality in the past 100 years ? Look around in your own house.

Look around in my house? Well my gf is at work so most fantasy land stuff is there!

I'm glad you think this project will happen. At least the main dev is accomplished.

But I hope you werent one of those guys that wanted an Amijoe, Boxer, Pios, or nAtami ;)