Support for PowerPC

[barracuda156]

FreeBSD supports ppc/ppc64, macOS supports ppc and partly ppc64.
Has something been done in this regard?

If it can run on both PowerMac G4/G5 and IBM Power-based machines, that would be a bomb.

[DKnoto]

Such topics don't have much appeal here ;)

IBM Power9/Talos II

[AssassinK786]

There is a PowerPC notebook: https://www.powerpc-notebook.org/en/

[MPC7500]

@AssassinK786 this is a pale horse

[AssassinK786]

@MPC7500 It is not pale horse,

To complete the schematics design, the designer will need around another two weeks.
From the beginning of August we already have in our hands a large part of the new design (Desktop version of Powerboard Tyche motherboard). We have started a fast review thanks to a few members of our association that were able to check the schematics. As usual in Italy, during August companies close for at least two weeks for vacations, so the designing is slowing down. For that reason, the PCB design will start after the middle of September.

We are proceeding faster than ever. We started the “Powerboard Tyche Desktop Electrical Schematics Design” donation campaign on 14th July, and the design itself started at the beginning of July. So after 1 month we already had a good part of the schematic done! Comparing that time with around 12 months that took the Notebook schematics design, now we are proceeding more than 4 times faster!!!

They have fast-forwarded their work and maybe at the end of this year, we may have the prototype properly tested.

Plus this is old development but recently a PPC board was designed and validated.

https://www.amiga-news.de/en/news/AN-2025-04-00129-EN.html
https://www.amiga-news.de/en/news/comments/329201.html

[barracuda156]

@AssassinK786 Waiting for it forever…

[AssassinK786]

@AssassinK786 Waiting for it forever…

Check out the other board.
They have mentioned on website

Yes, it’s possible to reach the goal to have a working Powerboard Tyche Desktop before the end of 2025, but is needed an extraordinary effort from donors because we depend on donations to cover all the steps that are coming : Schematics Design, PCB Design, Prototype Production and Tests.

If you can support through donations, then the process can speed up

[MPC7500]

This is really annoying now. You mention donations?
The project started in 2014. After years, they started to create a circuit diagram and PCB. For a notebook, to be precise.

Then they realized that the prototype didn't work. Now they want to create a desktop motherboard.
And they've already received tons of donations for the project they've been promoting for years...

I don't know what to call something like that.

Incidentally, this is a thing. They did it right:
https://mirari.vitasys.nl/

[barracuda156]

@MPC7500 It is great that they got the working board, but I am afraid performance will disappoint. 4 cores at 1.6 GHz without Altivec, from what it seems – it will be slower than 2005 PowerMacs.
Though if they make a laptop on that, that’s still interesting. (PowerBooks are beautiful, but slow.)

[MPC7500]

True. But to be honest, all Freescale CPUs are slow. The only advantages are power consumption and the integrated memory controller. It is more important to have a hopefully inexpensive PowerPC “bug detector” mainboard.

[AssassinK786]

This is really annoying now. You mention donations? The project started in 2014. After years, they started to create a circuit diagram and PCB. For a notebook, to be precise.

Then they realized that the prototype didn't work. Now they want to create a desktop motherboard. And they've already received tons of donations for the project they've been promoting for years...

I don't know what to call something like that.

Incidentally, this is a thing. They did it right: https://mirari.vitasys.nl/

I agree with your points.

1. The person who started this project is a non-tech person. If he were a tech person, this project would have been finished a long time ago.
2. Maybe they have failed on the notebook prototype, but they pivot very quickly to desktop.
3. The expertise required for this project was limited, as it worked as per donations received.
4. Raptor Computing Systems, which also builds OpenPOWER systems, is going to launch in Q1-26. Raptor Computing Systems anonymous tip on Talospace website & Raptor's Blackbird Arrives As The Most Open-Source Yet Fast Desktop System - Up to 8 POWER9 Cores, PCIe 4.0 dated 28 May 2019, took 6 years to create a new product, which shows that expertise comes at a cost.
5. The expertise is limited as people have moved from POWER to ARM, RISC-V, and x64 development.
6. After IBM, Google only came close to their expertise but went back to ARM + x64 servers and RISC-V because

They didn't provide much support to porting and optimizing SW for the Power ISA. IBM seems to have expected that interested users would port and optimize the SW themselves, because the POWER8 provided such a compelling performance benefit. That didn't happen. Users wanted a turnkey solution. IBM didn't provide one

So things will take time and money; it is up to us how to bring it to the mainstream, and I also appreciated this board. Incidentally, They did it right

[MPC7500]

What's annoying are your comments.
This project has been around for 11 years and you haven't gotten anything done. With the excuse that Roberto Innocenti isn't an engineer? Is Tim Cook an engineer? You have a relationship with ACube Systems and have hired someone to create the schematic and PCB.

And now you come here and contribute nothing to the actual topic. But you're making posts with false claims (#512 (comment)). Because there is no PowerPC notebook. Read the statutes of the project. It was exclusively about raising money for a notebook. As a donor, I could say I want my money back! I strongly assume that you are heavily involved in this project? Github is not a “beg for money” platform. That's pathetic.

BTW, I own a Blackbird myself.

[mszoek]

Hey guys - this issue isn't the place for such debates. Please take that elsewhere.

I have no objections to supporting ppc64 with ravynOS at some point. FreeBSD supports that platform and there were not massive changes needed in our code to support aarch64. I don't have the expertise in ppc64 assembly language do this though, so someone would probably have to volunteer to do and maintain the port.

[kevincrans]

For ppc support which has been removed after 10.7, this requires re-adding libraries.
We don't have to deal with 10.4 or Carbon as there were no 64-bit 10.4 or Carbon apps, so 10.5 and up, until we start talking about 32-bit & Nextstep.
There might be changes between Power4 and power6-8. Once all software works on the newest, will RavynOS run on the G5, and what about new apps?
That we're on Darwin means for more apps we either need BSD/Linuxulator back, some jit/bytecode or write them natively but that means porting Swift and Cocotron, and sticking to Power4 features or high requirements.
Lastly, multiple bootloaders should be supported, idk what the AmigaOne has, I also want to call the Xbox 360 and PS3 as possible targets.
For cross cpu-emulation and 32-bit one could create new issues.

Edit: I think without 32-bit support, it ain't really worth it.