dnetc benchmarks

[Iggy]

What I'd like to see, is how well Freescale's T4 and T5 AMP processor fair in a similar comparison.

Also, to be fair, the PA6T is a very nice Soc with great PCIe support.
No other PPC based processor offers as many lanes as this processor does.

[Iggy]

The poor performance of the PA6T could shed new light on the whole ISA shift affair by Apple. I mean, the PA6T was (inofficially) the scheduled successor of the 744x in Apple's laptops. But it performs rather worse. Not well suited for a 7447A successor laptop. ibm hat no fitting offerings for Apple and Freescale's 86xx were also not the huge step forward. No wonder why Apple looked elsewhere eventually.
But okay, Apple would have had one huge advantage over AOS4 and the X1000, tOS X has SMP support, so a Powerbook PA6T would have benefited quite a lot from the 2nd core.

Anyway, from those historical what-if games I would like to see an benchmark between a PA6T and 8641D based system. It seems to me as if Freescale's chip would have been the better choice eventually. A shame it was not used in a GP computer back when it was current.

I spent a lot of time working with the MPC8640/8641 (I still have two in my parts bins).
But as the older G4s clock as high or higher AND used Macs are cheap, the 86XXs didn't look like a viable alternative.
And the PA6T should be able to clock higher than either the G4s or the MPCs.

Right now, I'm looking forward to the T5040 (and, if offered, T5010 and T5020).

Not quite as nice PCIe support (as the PA6T), but it should perform slightly better.

[Iggy]

...there are no original MOS machines...

Pegasos I and II.
Perhaps you mean there is no new MorphOS specific machine available

...my Win 7/Linux on my quad core 4.8GHz pc is faster than your Linux box.

And my 3.6 GHz Quad core PC (while slower then yours) would also be faster, but that isn't the point.

[Iggy]

@Iggy

>>Pegasos I and II.
>>Perhaps you mean there is no new MorphOS specific machine available

Yes that is what I meant.

>>And my 3.6 GHz Quad core PC (while slower then yours) would also be faster, but that isn't the point.

It is exactly the point. We are comparing different products designed for different Operating Systems, and there are at least three threads which have been started trying to underline exactly what I said above.
So I am saying MOS is good and OS4 is crap, old Apple machine is better than new X1000, because this is exactly what these threads are about.

Ah, gotcha.
While I'm impressed with the X1000, I don't understand the logic in producing it.
Zylesea, myself, and other looked into current PPCs awhile ago and came to a similar conclusion.
While they're attractive products, the cost of producing a small run of PPC based systems is way too high (believe it or not, I don't think Aeon is overpricing their system - it really cost a lot to produce).

So when the MOS development team decided to support Macs, I shelved the PPC projects.

I doesn't make sense to rely on expensive custom hardware when existing low cost hardware (that can perform as well) is available.

The funny thing is, I was talking to Varisys about an MPC8641/8640 based board with an ATI SB600 Southbridge before it was announced that they had designed the X1000 (which uses the same Southbridge).

[Iggy]

Can you provide a link to that information? $500 for a CPU (particularly one designed for embedded systems) is farkin' ridiculous, that's what P3s cost new in 1999. Hope they didn't pay that kind of cash for all the parts...

I wouldn't know.

I got mine free
Freescale, like Motorola before them, has great developer support.
And I don't think that, in volume, the processors would be that expensive.

[Iggy]

Good points.

If I had any sense, I wouldn't have spent twice as much for my Mac CPU card as I did for my Phenom II X4 (which is more then twice as fast and has four times as many cores).

Who said we were sane?

If you can afford (and justify the purchase) of an X000, more power to you.

Somehow, I still feel more comfortable with that choice then AROS or WinUAE.

And, btw, you're not plugging a Radeon HD4650 into an old Mac (even if you could, there's no BIOS support).

[Iggy]

I quote from the exclusive interview with Trevor Dickinson from http://amigatronics.wordpress.com/2011/04/08/exclusive-to-amigatronics-28-questions-and-28-answers-with-trevor-dickinson-a-eon/

$500 for a PA6T.

Absolutely!
You're not going to get a good price buying a few hundred processors.

10,000, maybe.

[Iggy]

David, Daddy,
I wholly agree with you both.
I spent months poring over MPC8641, ULi 1541 and ATI SB600 data sheets.
I even started to enter initial information into some design software I have.

All the while I knew that any system based on what I was working on would be expensive to produce, difficult to sell and support, and have a very limited market.

And even after I'd given up on the idea, I found it tempting.

So I really DO understand Treavor.

And I admire Aeon's desicision on who to source the design to.

Varisys designers pointed me to the Qorlq line when I was still investigating the 86XX processors. They are some very knowledgeable people and I, for one, never doubted that the X1000 would get produced.

But I think we all have to admit this is a vanity project, not a sensible business venture (and I still envy them).

[Iggy]

Thank God this thread has gotten back to a more civil discourse (instead of the "mine's better than yours" rut).
Each NG OS and its hardware offers specific advantages relative to one another.
The fact that there are this many choices is a testament to how much interest remains in the Amiga platform.

[Iggy]

Wow!
takemehomegrandma.

Now that's a rant.

As I've said before, I don't think Treavor's investment was a sound business decision.
That said, I don't see the harm in it and I like the basic design of the X1000.
Is it too expensive? Obviously, if you're designing a new, low volume PPC based system that's inevitable.
If I had the resources, I'd be tempted to do something similar (although I'd use one of Varisys' recommended processors from Freescale, not a dead end like the PA6T).

Having someone build a vanity system based on idealized specifications (that will probably never be a commercial success) is hardly the disaster you paint it out to be.
Will it greatly increase the numbers of people using AOS4?
Nope.

Would I take one if offered?

So fast your head would spin.

[Iggy]

Good question, and given all the PPC suppliers are switching to ARM, one they are going to have to answer fairly soon.

Interesting opinion, but false.

Both AppliedMicro and Freescale continue to develop new PPC processors.

No current ARM processor outperforms even the dated G5.

Freescale has announce the re-introduction of AltiVec instructions to many of its PPC product lines and one processor that has as many as 20 virtual cores.

ARM is an in-order processor design with no where near the power of an out of order PPC design (per clock cycle a PPC is an significantly more powerful design).

All three of the most popular gaming platforms use PPC derived processors and at least two of their successors will continue to do so.

[Iggy]

I find it kind off odd (as a MorphOS) to find myself in the middle of this defending all NG OS'.

OK, so you don't like PPC. While a weaker performer then X86, PPC is clearly the most powerful alternate ISA. ARM may be a viable option soon. We'll have to see.

And I don't see the rush to move OS4 or MorphOS to X86. AROS already serves that market. And with OSX, Windows and Linux all going strong on X86 platforms, Amiga like OS' would be a hard sell.

BTW - Who really thinks there are blue and red camps? I just use the best alternative for me. For all I care, you can use whatever you want.

[Iggy]

Except OS4 doesn't support dual core.

Like that's a surprise?
No NG OS supports more then one core yet.

[Iggy]

Really?

Freescale - shipped low end ARM for years but have been recently been phasing out other architectures.  Yes, they are developing a high end PPC for networking but they have also licensed Cortex-A15 which will be similar performance wise.  Given the massive cost of developing CPU cores I can't see them doing anything other than die shrinks after this.

Applied Micro (AMCC) - They don't design PPC cores, they mainly license them from IBM.
They have announced a 64bit ARM for servers that is more aggressive and clocked higher than anything they have have PPC wise.

LSI - Another company doing ARM at the low end.  They had started doing PPC but they also just licensed the A15.

IBM - Go find me a PowerPC that you can actually buy as a chip that isn't a G3.
IBM don't make ARM chips but they are on Linaro's board - Linaro is a company that does ARM Linux.



A high end A15 should at least equal a high end G5.



Wrong. Both the A9 and A15 are out of order.

Wow, before you make claims you need to research your facts.

Yes both Freescale and Applied Micro have ARM licenses (and I believe even IBM has manufactured ARM processors under contract).

Applied Micro has downplayed all their former licensed designs in favor of several new PPC families that are considerably more powerfully.
Their 64 bit ARM processor is designed to be used in clusters in the server market (not for consumer products)

Freescale has launched the e5000 core (with 1 to 4 cores) and the e6500 core (in 4 core clusters with up to 24 virtual cores).
They are 64bit designs that are much more powerful then their predecessors.  
Hardly "die shrinks" by any stretch of the imagination (although they are slated to be manufactured on a 28nm process)

IBM continues to produce the Cell BE for Sony (PPC derived) and is designing the next processor for the successor to the Nintendo Wii that merges elements of PPC architecture with elements of the Power8 family.

IBM still offers G5 level processors (which are obviously more powerful then a G3), but no further development has been done of this line because there's no large buyers.
Instead, they've focused on Cell and Power derived solutions (which are both, again, PPC related).

ALL current (and near future) ARM processors other then AMCCs server line are 32bit (rumor is that Nvidia and Microsoft have also licensed with the intent to produce 64bit processors, but no official statements have been released).
ARM at its fastest runs at 2GHz (although some may soon make it to the "high" speed of 2.5GHz). The majority are in order execution designs.
The A15 will never best a 2.7 GHz G5 Mac and newer 64bit PPCs from Freescale should be able to do that.

Finally, let me make a personal prediction. Even if Microsoft is a licensee of 64bit ARM technology, the XBOX720 is still going to contain an INM designed PPC based processor (and ATI graphic). I'd bet on it.

ARM has an interesting future, but its not there yet.

[Iggy]

@ Iggy

A Cortex-A8 (Efika MX) beats a Sam440.
A Cortex-A9 beats a G4.

You like long winded posts, but you never offer any proof for your statements.

I'd really like to see a benchmark on  the second one since it seems implausible.

I've known about Nvidia's plans for some time, but the induction of Denver is a long way off. You could have just posted a wolf type link (rather then reproducing so much text).

And the XBOX720? So far, only news about the GPU (based on ATI HD 6000, now particularly impressive). Evolutionary not revolutionary. I'll take your bet.

ARM the future? Who knows?
I'm not sure I'd bet on anything (except seeing AMD continue to falter).
Against Intel, ARM has a real fight.