a) the Falcon is just as crippled as A600 and A1200.... 16bit bus on a 030?? WTF !!!
b) all P6T are 1.8 anything 2.0 is overclocked
c) learn proper quoting!!

>>vox wrote:

>Well I don`t believe them a bit. My chip is clocked at 1.8Ghz and originally it was introduced as 2Ghz only. Its now possible to buy it at 2Ghz at selected dealers.

To my understanding it is not possible to be bought as 2Ghz chip, because all chips ever manufactured were 1.8Ghz parts (run also at 2Ghz, though, but only as overclocked a little).

(you could sue those marketing those 2Ghz chips perhaps)

>>> OS4 doesn't support nearly half of board features now.
>>I think it supports more than half.
>Trust me, less then half. DOnt let me count em.

I try.

>CPU: PA Semi Dual-core PA6T-1682M, nominal 2.0GHz (1.8GHz standard) PowerISA™ v2.04+

Supported as well as existing OS can do it.

(Second core is not supported by AOS4 as AOS4 is not capable to do SMP yet. And IIRC, GCC for AOS4 does not yet fully optimize for PA6T.)

>4x DDR2 RAM slots
>10x USB 2.0
>1x Gigabit Ethernet
>2x PCIe x16 slots (1x16 or 2x8)
>2x PCIe x1 slots
>2x PCI legacy slots

All supported.

>2x RS232

I think they are supported.

>1x Xorro slot and xena

Supported in some ways (similarly as the FPGA of SAM?)
So far no developer has done anything signifficant with it and one can not blame a-eon for that I think.

Other supported/in use things are:
Onboard SATA2(not perfectly IIRC), audio,

GPU is poorly supported, only in 2D+compositing when AOS4 is used.
There is no 3D driver for modern GPUs on PPC amigalikes.

Not sure if onboard PATA is supported.

Anyway, it would be nice to see the lest of the another half of HW features that are not supported.

>>> A real DSP on board would be way more Amigish, like Falcon had.
>>Atari computers are "Amigish"? Interesting note
>Falcon is. more then crippled A600, A1200, A4000 and CD32.

LOL!
Btw PA6T has huge DSP performance via Altivec. And xena can do DSP functions also (just inferior when compared to Altivec capabilities).

xena is a bridge chip. One can use it to do things like bridge to A1200 trapdoor connector (like original eyetech/AInc idea was). (I'm sure no one will do that bridge) Another example would be to bridge legacy ports via xena and the available xorro prototyping board.

>> and adding more Xena cores (how when Xorro isn`t it? Via USB? PCI?)

To my understanding xcore links are available on the xorro slot. One should be able to add matrix of xcore chips there.

and...
>Cyrus was introduced rendering Nemo obsolete - correcting X1000 biggest gap and that is low CPU performance.

As long as AOS4 SW is not capable of using multiple cores, but capable of using Altivec of PA6T, Cyrus is not much faster.

>Sadly its again two versions of Cyrus with CPU soldered instead of one board with changeable CPU.

I think that would require a CPU card. And it would make Cyrus more expensive.
And there is three cyrus versions in the plans. x3500, x5000/20 and x5000/40.
(I think they should forget the x3500, though ... unless they manage to come close to SAM460 price)

IMO: x1000 is great HW. Only too expensive for me and for (too) many others.

[ Edited by KimmoK 04.03.2014 - 15:30 ]

Quote:

This keeps coming up, and I've yet to see a reason why this is something for which one would choose X-Core/Xena/Xorro (I'll just stick with 'X' from now on).
Sure, it's possible, but why is it desirable? In what way does this utilise X's 'special properties (low latency, multithreading, programmable, high bandwidth CPU connection)? How useful are these strengths in home automation (lighting, thermostatic, intruder detection, communication, audio/video switching, etc).

It really confuses me every time this is mentioned.

Sure, you can probably find something useful for it to do, but something that justifies its existence on the board in the first place?
Why is it a) there, and b) connected to the CPU like that. What practical application does that enrich above sticking it on a PCI card and calling it an 'option'?

Quote:

Well, yes 16 bit buts is a limit, not a first time for Atari,
but in its era, Falcon is advance over TT, while A4000 is
hardly an advance over A3000 if A3000 had AGA.


You are knowledgeable person, all PA Semis were 2Ghz
http://arstechnica.com/uncategorized/2007/02/8828/
http://www.pcworld.idg.com.au/article/171883/startup_launches_its_first_power-efficient_chip/
http://www.google.rs/url?sa=t&rct=j&q=&esrc=s&source=web&cd=3&ved=0CD8QFjAC&url=http%3A%2F%2Fwww.hotchips.org%2Fwp-content%2Fuploads%2Fhc_archives%2Fhc18%2F2_Mon%2FHC18.S2%2FHC18.S2T1.pdf&ei=HwYWU5bLLePD4gTP14H4Bg&usg=AFQjCNF_TJpe7b-6qxepLvp2gxiJVbS9MQ&sig2=AT8m9NA-A5VWPqcpgJYIhA&bvm=bv.62286460,d.bGE

Board could have better quoting system. But I can do old school >>> BBBS style.

Quote:

Quote:
And if pigs had wings...
The Falcon had plenty HW bodged together with 0 SW support, AGA in the A4000 was useable from day one, and a 040 was no CPU to frown upon at that time.
Quote:
Quote:

>> "G5" is usually considered as belonging to PPC970 microarchitecture family.
>> However, the PA6T is a unique microarchitecture of its own, developed from
>> scratch.

> True, but actually all G designations are Apple designations for chip generation

"G5" is the name Apple has used only for the PPC970 family, which the PA6T doesn't belong to (even though the PWRficient was originally intended to be for the Apple notebooks what the PPC970 was for the Apple desktops and workstations). So I wouldn't call the PA6T a "type of G5 CPU".

> Its kind of PentiumPro or 64-bit modern arhitecture designation.

Sorry, I don't understand this, and especially not as a reasoning as to why the PA6T should be called a "G5".

> Mint PPC G5 64-bit [...] does work on PA Semi.

The PPC64-compiled *userland* running on top of the Linux kernel works on any PPC64 CPU, but you need a special PA6T Linux kernel as the PPC970 Linux kernel doesn't run on the PA6T as is (nor on any other PPC64 CPU). That should tell you that the PA6T is no "G5".

> PA Semi is great chip

No, it was a company that was purchased by another company. The chip is called PWRficient PA6T-1682M (full name).

>>> It seems Pa Semi on Nemo is downclocked to 1.8GHz from 2Ghz

>> No, according to Varisys and A-Eon it isn't:
>> https://morph.zone/modules/newbb_plus/viewtopic.php?forum=3&topic_id=7183&start=472

> I don`t believe them a bit.

Fortunately what you do or don't believe doesn't change the facts. I for one trust Varisys in their statement, and EETimes agrees with that:

"Apple sent a letter to the DoD saying it will assure production of the 1.8 GHz PWRficient processor for three to five years, said one source who saw the letter but asked not to be named."
http://www.eetimes.com/document.asp?doc_id=1309557

> originally it was introduced as 2Ghz only.

No, the PA6T was introduced (read: announced) as running *up to* 2.0 GHz, but they didn't follow through on that announcement. That's what Varisys and A-Eon have said in the link I gave you, and what Kronos has told you right here in this thread.

> Its now possible to buy it at 2Ghz at selected dealers.
> http://forum.hyperion-entertainment.biz/viewtopic.php?f=33&t=2199

No, it's not. It's still running at 1.8 GHz. Trevor even clarifies that right in the thread you link to. May I propose that you go out of your way and read your own link?

>>> OS4 doesn't support nearly half of board features now.

>> I think it supports more than half.

> Trust me

I won't, that's for sure. You utter way too much false statements and also outright nonsense ad nauseam, even after confronted with evidence, for me to trust you.

> less then half. DOnt let me count em.

Please do count them. Else I won't believe you but instead maintain my opinion that OS4 supports more than half of the Nemo features.

>>> A real DSP on board would be way more Amigish, like Falcon had.

>> Atari computers are "Amigish"? Interesting note :-)

> Falcon is. more then crippled A600, A1200, A4000 and CD32.

Your statement that the Atari Falcon is more "Amigish" than Amiga 600, Amiga 1200, Amiga 4000 or Amiga CD32 is laughable at best. It may work if you define Atari as "Amiga" and Amiga as "non-Amiga", but why should anyone do such nonsense? Amigas are "Amigish" by definition, that's the meaning of the word. The Falcon is clearly "Atarish".

>>> they did false advertising with its comparison to Transputer

>> Let's see what "they" actually claim with regard to Transputer: [...]
>> I can't really see anything "false" in these claims.

> Don`t let me find all adverts from that time.

Please do, else I will continue to refute your claim of "false advertising with its comparison to Transputer".

>>> and adding more Xena cores (how when Xorro isn`t it? Via USB? PCI?)

>> http://amigakit.leamancomputing.com/catalog/product_info.php?products_id=1135

> I have wrote only review of Xena and X1000, thanks.

Sorry, I don't understand.

>>2x RS232
> I think they are supported. [...] Not sure if onboard PATA is supported.

They are both: http://amigaworld.net/modules/news/article.php?storyid=6345

> GPU is poorly supported

GPU is no Nemo board feature anyway.

> only in 2D+compositing when AOS4 is used.

You can use a fully supported Radeon card with R200 GPU in a PCI slot. It will be slowish due to 32-bit 33 MHz bus, but it will work.

> A4000 is hardly an advance over A3000 if A3000 had AGA.

But the A3000 doesn't have AGA, right? So what's the logical conclusion as per your own statement?

> all PA Semis were 2Ghz

No, the PA6T was not. It was and is a 1.8 GHz part which is possibly overclockable to 2.0 GHz.