Update:

>> 1.5 weeks ago A-Eon changed the X1000's official specs from
>> 1.8 GHz to 2.0 GHz.

> 3.5 weeks ago they changed it to "nominal 2.0GHz (1.8GHz standard)".
> http://www.a-eon.com/x1000.html

According to Paul Gentle of Varisys, the 2.0 GHz version of the PA6T was never on general sale (if it exists at all):

"The frequency that was quoted on the PA marketing collateral was never offered as a standard part option."
http://obligement.free.fr/articles_traduction/itwgentle_en.php

Edit: And now also Trevor Dickinson:

"despite the initial annoucement by PA Semi no 2.0 GHz version were ever produced."
http://amigaworld.net/modules/newbb/viewtopic.php?topic_id=35671&forum=34&start=160#665677

"The P.A. Semi CPU was originally designed for up to 2Ghz operation, but the company never manufactured a 2GHz certified version, before being purchased by Apple. So the CPUs on all NEMO boards are officially rated to 1.8GHz."
http://forum.hyperion-entertainment.biz/viewtopic.php?p=24961#p24961

[ Edited by Andreas_Wolf 30.01.2014 - 18:06 ]

Update:

> "The way in which it will be implemented however is already clearly defined
> and was subject to peer review by other developers. Obstacles to an efficient
> implementation were removed (e.g. the use of Forbid) and replaced in many
> OS components over the years (e.g. DOS).The foundation for SMP support was
> put in place, a clear picture exists what needs to be done to accomplish it and
> how. I'm willing to take a bet that it won't take 2 years ;)"
> http://amigaworld.net/modules/newbb/viewtopic.php?topic_id=34171&forum=33&start=40#627520

Reassuring words half a year later:

"The multicore support design for AmigaOS allows support of all of these features over time starting in AmigaOS 4.2."
http://amigaworld.net/modules/newbb/viewtopic.php?mode=viewtopic&topic_id=35285&forum=17#655578

Interesting. I wonder what they mean by "starting in AmigasOS 4.2".

> I wonder what they mean by "starting in AmigasOS 4.2".

I guess it means what it says, i.e. that OS4.2 is supposed to be the initial OS4 version to have a
"multicore support design" (whatever that means in particular). But yes, it's interesting that he is using present tense instead of future tense.

>"multicore support design"

That just makes my head hurt. Either it supports SMP or it doesn't.
That phrase doesn't tell you anything.

Don't spend too much effort on trying to decrypt these words, it's just the classic HyperionMP (AKA Ben Hermans) style. I will put these in the same category of the promises of protected memory support and resource tracking that date back to 2002. Of course, I'm always willing to be proven wrong, but until I see something concrete, these are just words from Ben Hermans and I know how much to value them...

Kind regards,
Andrea

[ Edited by guruman 06.03.2012 - 00:22 ]

Quote:

@boot_wb,

Just for entertainment, what is your estimate of how long it would take the MorphOS Dev. Team to complete support for the G5 PowerMac's, after they are finished with MorphOS3.0 for the G4 PowerBook?

Let's make it simpler and assume that the MorphOS Dev. Team has completed MorphOS3.1, or even 3.2 and all work on any of the devices contained within the G4 PowerBook's has all been completed, so no other work would be slowing them down from concentrating only on porting MorphOS3.x to the G5 PowerMac (no work on G5 iMac's yet).

My previous message sort of implied that I thought it would take them about 1 to 2 years to finish MorphOS3.0 and to complete the first release of any version of MorphOS which would support any of the G5 PowerMac's.

It is my opinion that the work to port to the G4 PowerBook and to implement any sort of wireless networking for MorphOS3.0, is probably much harder and more time consuming, than the work that will be required to port MorphOS to any G5 PowerMac. I could be completely wrong, because I don't really know anything about the differences, between G4 and G5 processors and how those differences will effect porting of MorphOS. The only other parts of the port that could slow it down and be time consuming, in my opinion, are the SATA controller and the power control & fan management (also liquid cooling pump control in the high end models) for keeping these high heat producing G5 CPU's from over-heating.

Other than those few devices, and probably a new Ethernet port to support, I don't see as many new things to work on, as there were when they started working on the G4 PowerBook, such as its CardBus/PCMCIA slot, new Radeon Mobility chip, battery charging/shut down/sleep, features, and many more.

If the team can finish MorphOS3.0 within the next 2 to 4 months (which I think is reasonable, considering that they thought they would be able to release it before the end of last year, and it is now two months later than that), I think it might take another 6 months to finish the rest of the features that the Dev. Team wants to support on the G4 PowerBook's. Then they should be free to start working on something new, like G5 PowerMac support (I hope).

Assuming all of the above, my guesstimate of how long it might take the MorphOS Dev. Team to complete a first release of a G5 supporting version of MorphOS, would be approx. 10 to 16 months after all work on the G4 PowerBook has stopped.

That gives me the impression that it might be possible for the MorphOS Dev. Team to complete support for an initial G5 support release of MorphOS by the end of 2013 (plus or minus 3 to 4 months.

Of course this is all wishful thinking on my part and I have no knowledge about anything that is happening within the MorphOS Dev. Team.

[ Edited by amigadave 05.03.2012 - 16:25 ]

Considering how quickly things have evolved since Mac compatibility was first introduced I'd say David's timeline is reasonable.
Some of the current work (like the R300 driver) moves straight to the G5 unchanged.

Hi guys, I remember a morphos team member explaining morphos3 will be the last version before a new and long way of dev to get a new System.
Are you sur G5 support is planned?

I don't know if this long hard work is to develope Qbox for a new system on PPC processors or if it's a new system based on Qbox for other processor family.

Don't you think G5 support should be wast of time?

I remember the gool of morphos on it's first times. It was to be a moderne operating system with SMP, resource tracking, memory protection and more based on Qbox and using Abox only for compatibility raisons.
Isn't it time to go this way?




[ Edited by serge 06.03.2012 - 07:42 ]

I still want an Lbox to run along side the Abox... especially if it ran on a different core than the Abox...

> Are you sur G5 support is planned?

No, hence the discussion in this thread on how long it will/would take to implement G5 support *if* it is/was planned.

@serge

Quote:

No.

And from some comments I have seen from MorphOS developer team members, it seems it won't happen.

Quote:

Yes, very much so.

Quote:

No, I don't.

@serge,
Quote:

No, I don't agree that it is time to go away from the PPC, not until all of the G5 PowerMac's and what ever few G5 iMac's that are practical and possible, are supported by a version of MorphOS that can run on them. Was this "explanation" that was given to you a private conversation, or something you heard with another group of MorphOS users, or read online somewhere? When did this explanation of what would come after MorphOS3.0 happen? Who was it that was giving this explanation to you?

I hope that this long hard work that you are writing about is some form of multi-processor support, instead of work to more to a different architecture. If OS4.x has figured out some way to provide multi-processor support into OS4.2 (which they are already talking about working in the present tense as linked to earlier in this thread), I am sure that the MorphOS Dev. Team can do the same kind of multi-processor support in MorphOS, and do it just as well, or even better.

The G5 is a significant jump forward for MorphOS that will give us much greater capabilities than ever seen before for any Amiga inspired PPC system. With some form of multi-processor support in MorphOS3.x, the dual G5 and dual core G5 models will bring even more power to MorphOS users.

Also, I thought that the MorphOS Dev. Team had given up on ever doing anything with the "Q-Box" and decided to instead just stick with working on the "A-Box"? If something has changed and the MorphOS Dev. Team has changed their minds about working on the "Q-Box", please send me a link to this new information, or statement by any of the MorphOS Dev. Team members.

It is my hope that the MorphOS Dev. Team will finish MorphOS3.0 and then continue to improve and perfect their support of the G4 PowerBook's. After that, I hope that they will begin work to turn their proof of concept for the G5 PowerMac's into full support for all G5 PowerMac's and also support for the few G5 iMac's that have video cards that can be supported by MorphOS.

After all work to support G5 Mac models is completed, if there is no other PPC designs that have been produced that make sense to port MorphOS to that offer better performance than the G5 Mac models, and also are available at a reasonable cost, then I think the MorphOS Dev. Team should merge with the AROS developers and create a combined AROS/MorphOS for X86, instead of re-inventing the wheel all over again on X86 by themselves.

[ Edited by amigadave 06.03.2012 - 08:17 ]

Hi Dave,

Enabling 'Wild speculation' mode..

Given that
- Efika took 1.5 years from announcement to initial release with all hardware documentation available (2 years until soundchip support);
- Mac support has been rolled out over >2 years; and
- Powerbook support has been around 2 years from Piru's 'teaser' video
I would estimate at least 2 years once porting work begins to an initial 32-bit/single-core/incomplete-drivers release, especially given that MorphOS team only release when there is a reasonably complete & stable support in place.
The memory controller/pci bridge is completely different chip to the G4 models, in addition to any on-board peripherals.

The development/release cycle (whilst fast for Amigaland) takes a while as evidenced by the current wait for 3.0. Since this was expected in December (as recently as October), and the delays seem to be down to a couple of key devs being out of commission, this highlights a vulnerability in MorphOS development: it's a spare-time development job for devs who are in demand in the salaried world.

It's all finger-in-the-air stuff, but I'd expect a two-or-three 3.x releases after 3.0 over the next 2 years adding further support for laptop peripherals/features. Following this, my crystal ball becomes less clear: G5 support? A shift in architecture (ARM please!)? Or a stalling of development/loss of inertia?
I suspect sales figures will have a large influence on this..
Currently MorphOS lacks any real strategy to attract new users. It's available for cheap hardware, but is it desirable for non-Amigans? How do people discover MorphOS? What features/software/advantage does MorphOS offer for 'ordinary users' who make up the bulk of any userbase?
These questions need addressing to inform any realistic development plan beyond appeasing the current userbase with faster hardware.

I'm with TMHG on this - why spend an additional 2 years supporting G5s, only to end up at the bleeding edge of a dead architecture (in terms of general purpose computers)? That time would imho be better spent porting MorphOS to a currently evolving architecture, ensuring a path for future development beyond 2016.

In summary, I wouldn't expect an initial G5 release for 4 years at least, if at all. But I'd love to be proved wrong...

amigadave,
Quote:
Considering that that (=Somewhat sensibly priced new PowerPC hardware) hasn't happened between 2007 and now, what makes you think that would happen any time soon?

Also remember it's not all about performance. Many would be happy with some "Sam440-grade" hardware, if the price reflected the specifications. A form factor is yet another thing. Personally I like small systems more than huge and noisy boxes like G5.

> (=Somewhat sensibly priced new PowerPC hardware) hasn't happened
> between 2007 and now

Actually, both the PowerStation and the LimePC series had their initial release in 2008, and the LimeBook was not released before 2009.

Quote:

please skip G5, runs too hot, too much power leeching.
let's go for modern and power efficient CPUs.

amigadave,
Quote:

The multi-processor support on OS4 planned is not SMP. Not real multi-processor support. I heard it long time ago but I dont think they changed this.
They want on a first time acces second CPU using a CPU context switch like we did on old days with Blizzard cards.
This way doing multi-processor support is crap.
But I'm agree with the fact if we cant have real SMP with OS4 or MorphOS on theyr actual form, so having this crap context CPU switch is better than nothing.

This limitation and lot of others are the reasons to wake up Qbox and begin to use it as main system. Abox should exists only as a compatibility layer and have to let place to Qbox.

G5 support have more than one problem.
G5 is not easy to support like other old PPC processors because it have some changes. controlers are diffrent to.
Porting MorphOS as it is today to G5 is too much work and starting the Qbox revolution is much work and lot of this work should go to trash if MorphOS Team decide to go to modern processors on futur because the Kernel will be modified a lot.

So, supporting G5 is IMHO a very big lost of time because if this work is done, it will not be used for futur versions.

serge,
Quote:

I am very much on your line. I also think the idea of the QBox should get reconsidered and could provide a way into an endianess idependent future. As far as I know most devs don't reject or dislike such an idea per se, but have their serious concerns about the required work that will be required. A helluvalot work! That's probably the main prob.
For completeness I add my view about the QBox and x86 (or ARM) for n-th time: http://via.i-networx.de/q86.htm

Since the future is uncertain anyway, I rather hope for the Powerbook release which will bring much fun and - hopefully -many new registrations and users.

boot_wb,
Quote:

Your guess is as good as mine. My shorter term guess is based on my belief that the two "Proof of Concept" demonstrations that were shown for the dual 2.7GHz G5 PowerMac, and the 2.1GHz G5 iMac, show that the at least one or two of the MorphOS Dev. Team members have an interest and an understanding of what it took to get MorphOS to boot on those two G5 models. This encourages me because if they can figure out how to do those presentations, I am hopeful that the amount of work to support the G5 is less than some people might think and that some of that work has already been at least partially completed, or the demonstrations could never have happened. As another member here pointed out, all of the work on video card drivers, will transfer directly to support for those same video cards being used in the G5 PowerMac's.

What it really boils down to is what the MorphOS Dev. Team wants to support next. Only those Team members can really know how much work is already done, from when the demonstrations were presented.

I was not at any of the shows where Big Foot, or other Team members demonstrated MorphOS booting on any G5 system, so I can't say how much, or how little work was completed,just to get it booting.

Did they boot from a USB stick, a CD, or from a SATA hdd?

It could be that much less work was needed to do those "Proof of Concept" demonstrations than I had hoped was already done. Just more of my wishful thinking that more is done and less needs to be completed, so G5 support could be provided for MorphOS.

I still think that it would be better for everyone, if the MorphOS Dev. Team decides to move to x86 and/or ARM, that they do it with the AROS developers, instead of in competition with them. They can even do it alongside them with their own MorphOS fork of AROS. To start over from scratch with a new x86 and/or ARM creation would be a mistake, in my opinion.

I do understand why many MorphOS users want MorphOS to move to x86 and/or ARM now, instead of later.

@serge,

I understand that any implementation of multi-processor support for OS4.x, or MorphOS3.x, would not be true SMP and therefore might actually be crap. But as you wrote, it might be better than nothing, if it can provide a small boost in performance, or usability on systems with dual G4's. It might not be worth the effort to even create it, but since Hyperion have virtually promised to provide it to OS4.x users, I am sure they will complete something and we can then evaluate if it is useful at all, or complete waste of time and effort. We will have to wait and see.

The context switching was a killer on the slow Phase5 boards of the past (slow compared to what is available today). Now with today's modern components and CPU's that are orders of magnitude faster than those on the Phase5 boards, maybe the context switching won't be as noticeable at slowing down the system, compared to the added processing power a second core of the 1.8GHz PA6T can provide (or a second 1.42GHz to 1.8GHz G4 could provide on a MorphOS3.x system, if the same kind of AMP were implemented. The MorphOS Dev. Team can wait and see if what Hyperion comes up with is worth the effort, before they decide to copy their efforts, or come up with a better AMP solution).

Other members here are probably right and the MorphOS Dev. Team will choose to move to x86 and/or ARM, instead of wasting any more time and effort on PPC, or less than perfect AMP strategies.

I just wonder how "Amiga" like they will be able to keep MorphOS after such a move. The "Legacy" layer that provides backward compatibility might be like a virtual machine, or some kind of seamless integration of UAE to make it invisible to the user.

The most important question is how can the MorphOS Dev. Team find more users and developers? They have not done much in the past, as MorphOS has not really been ready for non-Amiga users. I think it is polished enough to compete with Linux for people that are willing to put up with some missing features and who are interested in different operating systems.

[ Edited by amigadave 07.03.2012 - 11:54 ]

> The multi-processor support on OS4 planned is not SMP.

Hyperionmp contradicts you:

https://morph.zone/modules/newbb_plus/viewtopic.php?forum=3&topic_id=7183&start=336

> I heard it long time ago but I dont think they changed this.

Hyperionmp's positive statement on SMP is from half a year ago.

> Did they boot from a USB stick, a CD, or from a SATA hdd?

https://morph.zone/modules/newbb_plus/viewtopic.php?forum=11&topic_id=8313&start=23

> today's modern components and CPU's that are orders of magnitude faster
> than those on the Phase5 boards

"Orders of magnitude faster" means at least 100 times faster, which would equal a 20 GHz "G2" CPU.

> or a second 1.42GHz to 1.8GHz G4 could provide on a MorphOS3.x system

...or even a second 2.0 GHz G4.

Quote:

OTOH, "Hyperionmp" is full of shit, and has been so for more than a decade...

>>> The multi-processor support on OS4 planned is not SMP.
>>> I heard it long time ago but I dont think they changed this.

>> Hyperionmp contradicts you:
>>
>> Hyperionmp's positive statement on SMP is from half a year ago.

> OTOH, "Hyperionmp" is full of shit, and has been so for more than a decade...

The disagreement between serge and me is about what Hyperionmp *claimed*, not about what is or will be for real.