> we "just need" POWER7 [...] support inna MOS 4.x PPC

POWER7 is not really suited for the ordinary home for several reasons.

New member:

http://www-03.ibm.com/press/us/en/pressrelease/43195.wss

Silently revealed already a week earlier:

http://www.hpcadvisorycouncil.com/events/2014/stanford-workshop/presentations/day2/7_IBM_Open_Power.pdf (page 4)

Oh yeah, Power8 is beginning to look promising.
I wonder if there will be affordable processors in the line.

Quote:

I wouldnt count on it since the only thing using them is hight end servers, and very few at that.

Quote:

Yes, but on the flip side, the Chinese are buying into this and they don't usually do high end projects.

If IBM wouldnt offer the level of customization required by Apple and caved to the video game console market in favor of lower margins in the millions of units sold, then I wouldnt count on anyone ever using them again in a consumer level.

First off not a single consumer level OS is currently supported, secondly no software is being made on that same level. They blew it with Apple and sold out to the video game console manufacturers. The two largest of which are now back on X86 class chips. Its not going to happen.

> [...] video game console manufacturers. The two largest of which are now back on
> X86 class chips.

It's not "back" for the PlayStation. PS1 and PS2 were MIPS, and PS3 was PPC. No x86 in this lineage before the PS4.

And this is a purely economic phenomenon.
The development of cheap APUs helped foster this.
Curiously, X86 designs began to look like the XBOX360's combined cpu/gpu core.

Excuse me, but what is the point of PPC in 2014, really?

There are no consumer products anymore that uses them since both the Playstation and X-Box have abandoned it. This means that no matter what anyone do, any consumer computer using PPC will be very very expensive. Maybe price isn't a problem for high end servers and such, but today when you can get a nice enough intel computer for a couple of hundred euros/dollars, why would anyone want to invest in a computer where the CPU alone costs several hundred euros/dollars (except for a handful of rich computer nerds)?

Quote:

Several hundred?
The one I'm looking at starts at around $150.
What's the point?
Simple, its compatible with our current ISA.
And there is no shortage of OS' for X86 or ARM, PPC for the end user is now pretty much ours.

Still pretty expensive. And it wont help if we take the A-Eon ruite and make our own mother board for a handful of users. The beta board testing cost would be enormous, which is why the X1000 is so expensive.

Then it makes more sense to "just" port MOS to the X1000 and X5000. But that still makes little sense to me.

An ISA shift in the long run to commonly used hardware is IMHO the only way to go.

> The beta board testing cost would be enormous, which is why the X1000 is so expensive.

In board development, the bigger chunk of the total costs is usually spent before the beta testing phase starts.

Quote:

Absolutely, just the calculations necessary for every trace get pretty intense.
Design is definitely the most rigorous part.

And Yasu, as the X5000 is not available yet, what makes you think it would not be supported?
This has always been a possibility.
Look at the current work being done on the SAM460 port (and those machines are relatively hard to get).

@Andreas_Wolf

You know that's what I meant

@Jim

Pure assumption since the X1000 isn't supported. My guess is that the MOS Team doesn't see much point either or just don't want too (so much work). Buying a G5 gives you much more bang per buck.

[ Edited by Yasu 17.02.2014 - 16:31 ]

Quote:

Yes, buying a G5 is a better economic solution, but they are not new.
Personally, I already have a PCI-E 2.3 GHz Dual Core G5 waiting in the hope that one day we will see support for those models (most of the bang of an X1000, with a faster processor).

But I'd consider buying an X5000 if support.
If only to say I had MorphOS running on an AmigaOne.

>>> The beta board testing cost would be enormous, which is why the X1000 is so expensive.

>> In board development, the bigger chunk of the total costs is usually spent before the beta
>> testing phase starts.

> You know that's what I meant :-)

No, I don't :-)

>> as the X5000 is not available yet, what makes you think it would not be supported?

> Pure assumption since the X1000 isn't supported.

At least the X1000's Hermans factor is missing with the successors, but that's not to say this alone would create enough incentive of course.

> Buying a G5 gives you much more bang per buck.

As Jim has already indicated, that's also true for the Sam460 yet MorphOS is in the process of being ported to it.

> If only to say I had MorphOS running on an AmigaOne.

Then buy a Sam460 (aka AmigaOne 500) with MorphOS and save expenses :-)

Quote:

I keep coming back to that idea myself quite frequently.
And I already have an X1300 video card.
If we don't see X5000 support, I may do just that.
It can replace my PowerMac G4's as my "low end" system, while I keep the G5s for more demanding applications.
If I am only using Odyssey, I hardly need G5 level performance AND the high power draw

Yeah, that would fit in nicely, Sam460, iBook, and PowerMac G5.

Silver members revealed:

http://www.open-power.org

This one among them:

"The company intends to extend its product line by developing future OpenPOWER-compliant systems based on IBM’s POWER microprocessors"
http://servergy.com/servergy-joins-the-openpower-foundation/


Edit: Gold members revealed:

http://openpowerfoundation.org/membership/current-members/

[ Edited by Andreas_Wolf 05.04.2014 - 01:27 ]

Xilinx two steps down from Altera.
Interesting.

@ Gold - PowerCore, I don't recognize that one.
Any info?

> PowerCore, I don't recognize that one. Any info?

See posting #14 in this thread. You even replied to that.

China Core, yeah I remember them.
In fact, I suspect a Taiwanese ARM licensee of slipping that group information.
They are actually one of the Gold member that might run with this, they have previous RISC experience.

Oh, and before we hear a repetition of the "what relevance does Power technology have" comments, its fresh and its not X86.

> IBM Opens Up POWER Architecture For Licensing
> The article mentions Nvidia favouring this development. Fair enough, but why?

"NVIDIA today announced that it plans to integrate a high-speed interconnect, called NVIDIA NVLink, into its future GPUs, enabling GPUs and CPUs to share data five to 12 times faster than they can today. [...] The new interconnect was co-developed with IBM, which is incorporating it in future versions of its POWER CPUs. [...] "NVLink enables fast data exchange between CPU and GPU, thereby improving data throughput through the computing system and overcoming a key bottleneck for accelerated computing today," said Bradley McCredie, vice president and IBM Fellow at IBM. "NVLink makes it easier for developers to modify high-performance and data analytics applications to take advantage of accelerated CPU-GPU systems. We think this technology represents another significant contribution to our OpenPOWER ecosystem." With NVLink technology tightly coupling IBM POWER CPUs with NVIDIA Tesla GPUs, the POWER data center ecosystem will be able to fully leverage GPU acceleration for a diverse set of applications, such as high performance computing, data analytics and machine learning. [...] Today's GPUs are connected to x86-based CPUs through the PCI Express (PCIe) interface, which limits the GPU's ability to access the CPU memory system and is four- to five-times slower than typical CPU memory systems. PCIe is an even greater bottleneck between the GPU and IBM POWER CPUs, which have more bandwidth than x86 CPUs. As the NVLink interface will match the bandwidth of typical CPU memory systems, it will enable GPUs to access CPU memory at its full bandwidth. "
http://nvidianews.nvidia.com/Releases/NVIDIA-Launches-World-s-First-High-Speed-GPU-Interconnect-Helping-Pave-the-Way-to-Exascale-Computin-ad6.aspx

It will be interesting to see how NVLink relates to POWER8's CAPI, which "is an overlay that will ride atop the PCI-Express 3.0 mechanicals to provide coherent memory addressing for CPUs and external coprocessors like Nvidia GPUs".

Nice, now if you were going to build a new age Amiga and you wanted a truly different approach...
Hmm, the time to start planning would be right about now.
Processor and Graphics? Well Andreas has pointed to one idea.
But what kind of I/O and Audio systems would you want?

It would be nice if they where to produce some computers that may be a little expensive but passable for personal computers. Maybe that really could be a chance for MorphOS. There aren't that many non-unix(like) systems out there for PPC.