> some apes began to throw shit on [...] powerpc technology in general. [...]
> Dear Jim [...]

You sound confused :-)

Comment from Hyperion .

http://amigaworld.net/modules/newbb/viewtopic.php?mode=viewtopic&topic_id=40622&forum=2&start=460&viewmode=flat&order=0#770947

[ Edited by Spectre660 17.10.2015 - 11:23 ]

http://www.os4welt.de/viewtopic.php?f=4&t=1670&sid=e2c6fbaecf723b05ca59e3531417292b

So, I was right about apes throwing shit on a-eon and powerpc technology in general.

Some people simply don't like bitter truth.

btw netbsd has port for e500v2 for three yrs, and users of this port also report

that trapping fpu instructions on e500v2 is not a problem, and has no impact on performance.

And where does who state that it would have no impact on performance ?

Trapping allways has a negative impact and replacing the FPU commands with integer code (the FPU on chip uses diffent datatypes and is unusable for emulation) will also slow things down.

How big that impact is will hugely depend on the app used.


So OS4 can handle that SoC just as good as AmigaOS could handle an 68LC60, only that back than trapping didn't work (as LCs were really downlabeled 68060s with a defective FPU).


Considering that FreeScale has SoCs with a proper FPU and similar price and performance on sale one can only wonder wether "Amiga" was even a faint 3rd though when designing that board.

Quote:
The post you linked doesn't say anything about no impact on the performance. Even Ben said no significant impact, but didn't back it up with anything. I prefer to be a thinking ape, rather than taking things at blind faith.

> netbsd has port for e500v2 for three yrs

You can't compare *BSD or Linux with OS4 as the latter doesn't get all 3rd party programs recompiled with each OS release.

> users of this port also report that trapping fpu instructions on e500v2 [...]
> has no impact on performance.

Do you have such link where performance of FPU-depending binaries on e500v2 is compared between FPU trapping and compilation for e500v2 and no difference is noticed?

Quote:

The Grape Koolaid is strong in this one. So much stupid, so little time.

Quote:

A comment from Ben Hermans isn't worth the electricity used to light up the pixels that make the words of his statement. This man has a history of making laughably false statements. Not just statements from 10-15 years ago, but recent history too.

> the FPU on chip uses diffent datatypes

Both the standard PPC FPU and the e500v2 FPU can handle double precision floats. Or what do you mean?

> FreeScale has SoCs with a proper FPU and similar price and performance on sale

True, but I suspect that QorIQ T1 wasn't available when Tabor idea was born and design process started. Nevertheless, they could have used a Mamba SoC from APM's PPC465-based PacketPro family instead.

http://www.apm.com/products/embedded/mamba-family/

Quote:

Exactly. Lawyer and liar sound similar and _sometimes_ are similar.

The bitter truth? We are not really talking about a revolutionary piece of hardware here, even if we ignore the FPU.

My cheap Mac Mini will still run MorphOS faster than AmigaOS 4 will run on that board. _That_ is the bitter truth.

Quote:

I wonder how it'll compare performance wise to OS4 in UAE on an identically priced x86 machine?

Quote:

Pretty well I'd say. Mac Minis are still cheap as dirt, and OS4 Classic have only very limited hardware support (128MB RAM, no HW graphics acceleration, etc.).

Quote:

Sorry, I should have been more specific, I meant "I wonder how the Tabor will compare" not the Mac Mini.

If the Tabor sells for 700 you can buy an awful lot of Intel brute-force emulation power for the same price. UAE will probably get hardware accelerated graphics support and more emulated RAM at some future point too I imagine.

The limitation for OS4 emulated lay in OS4-classic which will bail on anything that goes beyond what the actual HW could deliever.

"Fixing" OS4 to support UAE more directly would be trivial, but just won't happen.

Porting OS4 to a no-nonsense PPC-EMU would also be possible but won't happen till Mr Beelzebub puts warm socks on his XMas wishlist.

Quote:
I think that was understood. Still, given that this is a MorphOS website, it should not be surprising that someone would point out how a low-cost Mac Mini G4 would provide much better performance than PPC emulation on a very high-end Intel-compatible machine.

Quote:
Single-core performance has not improved much in the past few years. Intel's focus appears to have been on improving power efficiency.

Quote:
We'll see.

Quote:

I talk with few users, and they say that it is not a problem.

aka, they do have a rather odd definition of "problem".

Sure stuff will run, and if it ain't performance critical (and noone sane would use PPC-Linux when performance is a concern) one would probraly not notice the difference.

Doesn't mean there ain't no one....

Yes, apps works without bugs and visible slowdowns,
it means there is no problem with fpu trapping.

Quote:

What is a "visible slowdown"? Where are the benchmarks running on OS4 or at least linux?

I suggest that all speculation be delayed regarding performance and pricing of the AmigaOne A1222 until after benchmarks are available, and A-Eon has finalized the price which they are going to charge for it. Take the statement from Trevor who has said that the 700 to 1,000 Euro quoted price is "very high" over what the actual price will be, as proof that he is doing his best with A-Eon in partnership with ACube to bring a lower priced new AmigaOne system to the users.

As for the FPU complications and performance implications, I'll wait for actual benchmarks on real world applications and games before passing judgement, as the developers who are actually doing the work of porting to this board are showing very little concern that the performance will be unacceptable. If it were truly that bad, I doubt that they would spend their time working to port to this product.

In other words, I think that most posters in this thread need to calm down and take a "wait and see" attitude, before claiming that the sky is falling.

[ Edited by amigadave 18.10.2015 - 10:34 ]

The sky is not falling. We already have good range of supported hardware, old and new, slow and fast, cheap and expensive, portable and desktops. Tabor motherboard offer nothing new to us.

Quote:

We will see if the MorphOS Dev. Team decides if the A1222 is worth porting MorphOS to or not. I don't think it will be a system I am personally interested in buying, but there are many AmigaOS4.x users who have been begging for a cheaper system to purchase to run AmigaOS4.1FE on, and there might be a few MorphOS users who also want a new system that is lower priced than the X5000.

I am not advocating a port to the A1222, but I am also not worried, nor will I complain if the MorphOS Dev. Team decides to support the A1222 by porting MorphOS to it. As has been stated several times already, the port of MorphOS to X64 or ARM hardware is going to take a long time.

Bigfoot told me at Neuss that they are not at all interested at porting MorphOS to this board and mentioned the incompatible FPU as one of the reasons.

@ Yasu

Frank Mariak posted a similar comment here on MorphZone the other day:
MorphZone Post