Quote:

There's nothing wrong with the chips or the boards but in the minute volumes you have you there will be no alternative but to pay X1000 style prices for them. I'd say that's a pretty big reason.

Do you really want to pay top dollar for a machine which will be beaten by a phone in the next couple of years? Then Indian $35 tablet will soon have better hardware!


Changing platform is an opportunity. It's going to break compatibility so why not do it properly and add all the stuff that's been missing - (memory protection, SMP etc, etc...).

It'll take time, but changing platform will give you cheap, powerful hardware *and* a potentially much upgraded OS.

[ Edited by minator 12.04.2012 - 22:40 ]

minator,
Quote:

You might as well just switch to Linux or AROS now and modify that to look and act more like MorphOS. It seems to have what you are looking for.

I would not call my dual 2.7GHz G5 hardware comparable to a cell phone and it also did not cost in the price range of the X1000. I can find a dual 2.7GHz G5 PowerMac for you too, probably for about $250, maybe less.

I am not opposed to MorphOS eventually moving to another platform, but I would prefer if they move to x86 that they would join forces with the existing AROS developers. I would also prefer that they complete the port to the 2.7GHz G5 PowerMac, just to show how fast and powerful MorphOS on PPC could be, before they make a change to a different platform.

It is all up to the MorphOS Dev. Team, so it does not matter what anyone else wants, as you and I do not have a vote in the matter.

minitor wrote:
Quote:

How long will it take to port MorphOS to x86 and how different will it be by the time it is done? It looks like the MorphOS user base shrank by half or more when they moved from being free to the paid for version 2.0. I would guess that it would probably shrink by half or more again, if it takes 18 months to 2 years, before a beta version of MorphOS on x86 was ready to release and it is little more than a commercial version of AROS with a MorphOS look & feel.

[ Edited by amigadave 12.04.2012 - 18:01 ]

Quote:

Same here!

MorphOS on x86???

Do not know.... hmmm....

>> I'd like this in a 2.0ghz setup.

> Same here!

Then you will have to wait for the T5 as the T4 has been relegated from 2.0 to 1.8 GHz.

How many PCI lanes does the T4 support?

> How many PCI lanes does the T4 support?

AFAIK information about the number of PCIe (note the "e", legacy PCI is not supported) lanes is not public yet. What's been revealed is the number of SerDes lanes, of which some will be configurable as PCIe lanes (16 at best in case of T4240, 12 at best in case of T4160; but due to SATA controllers will have to be decreased by 1 or 2 in both cases), and the type and number of the PCIe controllers:

https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=7901&forum=11&start=8

> these threads are utterly pointless and a waste of bandwidth

...whereas those threads are useful and an exploitation of bandwidth:

https://community.freescale.com/community/the-embedded-beat/blog/2012/05/03/expert-viewpoint-fused-core-logic-multithreading-for-multicore-performance

;-)

Addendum:

>> How many PCI lanes does the T4 support?

> AFAIK information about the number of PCIe (note the "e", legacy PCI is not supported)
> lanes is not public yet. What's been revealed is the number of SerDes lanes, of which
> some will be configurable as PCIe lanes (16 at best in case of T4240, 12 at best in
> case of T4160; but due to SATA controllers will have to be decreased by 1 or 2
> in both cases), and the type and number of the PCIe controllers

Looking at http://www.ces.ch/images/stories/ces/product_announcements/riov-2440-product-news.pdf it seems that 16 PCIe 3.0 lanes are possible with the T4240, and that's even (contrary to my expectation) with the 2 SATA2 controllers being active at the same time.

Update:

> "Availability of the royalty-free Mentor Embedded Performance Library for Freescale's
> AltiVec technology is planned to coincide with initial availability of QorIQ AMP silicon."
> http://media.freescale.com/phoenix.zhtml?c=196520&p=irol-newsArticle&ID=1603798

New AltiVec libs now available:

http://media.freescale.com/phoenix.zhtml?c=196520&p=irol-newsArticle&ID=1830586
http://www.freescale.com/files/32bit/doc/app_note/AN4786.pdf


Edit: added PDF link

[ Edited by Andreas_Wolf 12.09.2013 - 03:27 ]

> "AltiVec technology is planned to coincide with initial availability of QorIQ AMP silicon."

Odd, unless someone is getting pre-production silicon, they released this before they said they would.

What does all this have to do with MOS?

bash64,
Quote:

If you don't know, then explain it would only make you look foolish.
Basically, MorphOS' ISA is PPC.
This is one of the most powerful PPCs yet to be released.

Quote:

Nothing at all!

Quote:

MorphOS runs on PPPC CPUs, there's nothing wrong in a discussion about the current state of PPC hardware, i find it interesting to read these threads, Andreas, Jim and a few others are discussing stuff i would have probably never heard about if they didn't, it satisfies my curiosity on the topic, and this is quality discussion they have.

Had to check and make sure I hadn't wandered onto a red forum by mistake...

SoundSquare,
Quote:

+1

>> "AltiVec technology is planned to coincide with initial availability of QorIQ AMP silicon."

> Odd, unless someone is getting pre-production silicon, they released this before
> they said they would.

Seeing as the T4240QDS has been available for about half a year, I guess that initial (i.e. pre-production) T4240 silicon has been available for at least as long.

> What does all this have to do with MOS?

Nothing. You may want to ask the thread opener why he chose to start this thread here in "Hardware" instead of "General Discussion".

>Seeing as the T4240QDS has been available for about half a year, I guess that initial (i.e. pre-production) T4240 silicon has been available for at least as long.

I wasn't aware of that (obviously).

Andreas_Wolf,
Quote:
Well, the e6500 core is/will be part of quite some hardware. Whether that hardware will be a target hardware for MorphOS or not, doesn't cange a thing it's about hardware.

Plus, while I am generally voting for a switch to x64, MorphOS currently uses the ppc ISA and hence every processor of that ISA is somehow related to MorphOS IMHO.

>> Seeing as the T4240QDS has been available for about half a year, I guess that
>> initial (i.e. pre-production) T4240 silicon has been available for at least as long.

> I wasn't aware of that (obviously).

I think you were:

https://morph.zone/modules/newbb_plus/viewtopic.php?forum=11&topic_id=7780&start=21

Yes, my recall is suffering as today was my first anniversary away from my wife.
Good thing I have you as a backup.

> the e6500 core is/will be part of quite some hardware. Whether that hardware will
> be a target hardware for MorphOS or not, doesn't cange a thing it's about hardware.
> [...] MorphOS currently uses the ppc ISA and hence every processor of that ISA is
> somehow related to MorphOS IMHO.

"The Hardware forum is utilized to provide a place to discuss hardware related issues involving MorphOS." may be interpretable as meaning that MorphOS must already run on the hardware. Personally, I couldn't care less in what subforum this thread resides but obviously some people think it's out of place here, and judging by the description of this subforum they may have a point. Needless to say that the final decision is up to the mods/admins, and as this thread has been living here for the past 2 years it's highly probable that this is indeed regarded as the right subforum by the powers that be.

Update:

> I guess that initial (i.e. pre-production) T4240 silicon has been available for at least as long.

Prices are now public:

Freescale:
T4240 1.67 GHz: 570 USD (1000 pcs. qty.)
T4160 1.67 GHz: 462 USD (1000 pcs. qty.)

Future Electronics:
T4240 1.67 GHz: 800 USD (single qty.)

DigiKey:
T4240 1.8 GHz: 833...866 USD (from 25 pcs. qty. to single qty.)

Avnet:
T4240 ??? GHz: 835...916 USD (from 50 pcs. qty. to single qty.)


Edit: updated prices

[ Edited by Andreas_Wolf 29.06.2014 - 23:29 ]

Update:

>> How many PCI lanes does the T4 support?

> AFAIK information about the number of PCIe (note the "e", legacy PCI is not supported)
> lanes is not public yet. What's been revealed is the number of SerDes lanes, of which
> some will be configurable as PCIe lanes (16 at best in case of T4240, 12 at best in
> case of T4160; but due to SATA controllers will have to be decreased by 1 or 2 in both
> cases), and the type and number of the PCIe controllers

According to http://www.freescale.com/files/training/doc/dwf/DWF13_APF_NET_T1284.pdf (page 16), it's indeed possible with the T4240 to have up to 16 SerDes lanes for PCIe using either 2 controllers (x8 x8), 3 controllers (x8 x4 x4) or 4 controllers (x4 x4 x4 x4), whereas each controller can be downgraded to use less lanes. Those configurations would be without SATA though.
Using 2 lanes for SATA leaves 14 lanes for PCIe, which can be configured as using either 3 controllers (x8 x4 x2) or 4 controllers (x4 x4 x4 x2).
While two out of the T4's four PCIe controllers support PCIe 3 standard, the following restriction is mentioned by Freescale:

"8 Gbaud is not supported with x8 link width"
http://www.freescale.com/files/training/doc/dwf/DWF13_Designing_QorIQ_Hardware_SanJose.pdf (page 51)

This means that whenever a PCIe3-capable controller of the T4 is being assigned 8 lanes, it switches from PCIe 3 to PCIe 2 standard, reducing the bandwidth of those lanes to about a half.

Edit:
http://www.freescale.com/files/training/doc/ftf/2014/FTF-NET-F0031.pdf has some more info on page 31/32:

Using 16 SerDes lanes for PCIe, the following configurations are possible:

1. x8-gen2 x8-gen2
2. x8-gen2 x4-gen3 x4-gen2
3. x4-gen3 x4-gen2 x4-gen2 x4-gen2