@pavlor

Aros offers this... it depends where you run it

@cdimauro

Yeah, I was trying to be nice but you're right. Any idea on what kind of performance we can expect out of the x5000? I suspect even its numbers will be somewhere in the middle of the past decade, taking into account the x1000's track record.

@ferrels

Hope so, then it would be the fastet amiga ever.

@eder


That's a good one eder! Hoping for mediocrity!

That one gave me a laugh......The fastest Model-T ever produced is still a Model-T! The fastest Amiga I've ever seen is a Classic Amiga emulated via WinUAE or Amithlon or the Vampire 600.

I still don't understand this obsession with decades old PPC processors that are mainly used to emulate 68k apps via OS4. If OS4 had some worthwhile apps, then I could see why, but there just isn't any compelling reason to pay thousands for an outdated, feature-lacking OS to emulate 68k apps on outdated hardware. Mental illness maybe?

Last edited by ferrels on 13-Jun-2015 at 12:35 AM.
Last edited by ferrels on 13-Jun-2015 at 12:34 AM.
Last edited by ferrels on 13-Jun-2015 at 12:33 AM.

10 years too late even IF they get it working. All those suckers got shafted who bought an x1000 to run ppc linux if they want to use both cores.

Did they ever even get all the motherboard drivers yet?

I guess 3,000 does not include drivers, maybe that costs extra.

3,000$ only gets you access to one core in amiga os. What a joke.

@haywiresystem

While I don't disagree with the sentiment that 3000$ is far to much for a AOS4 system, I do find the whole "AOS/Hardware" bashing somewhat funny.

AOS, MorphOS and AROS are all underperforming OSs that are a "waste of time and money". Yes, AOS is the OS that takes the biggest chunk of money out of your pocket, but for the most part, time is money.
I guess I should spend time on AROS and MorphOS forum ridiculing the users for "wasting their time" on a deadend OS project, and then throw a few "lololol" at the end of it.
Same goes for C64 users and classic Amiga hardware tinkers.

While the X1000 is very expensive for what you currently get out of it performancewise, its not just a hardware issue. As you alluded to, the driver issue seems to be the main hurdle at the moment.

I guess my point is; people tinker with "deadend" hardware and OS's (AOS, AROS and MorphOS) for the fun or challenge of it.

Last edited by Overflow on 13-Jun-2015 at 08:45 AM.

@ferrels

Quote:

what kind of performance we can expect out of the x5000?

Assuming DMIPS numbers published by Freescale are comparable across e500/e5500/e6500 cores, e5500 core 2.0 GHz is as fast as e6500 core (single thread) 1.8 GHz. T4240 1666 MHz scores 6.86 CINT2006 (base), so P5020 could score 7.4 CINT2006 (single core performance).

Quote:

I still don't understand this obsession with decades old PPC processors that are mainly used to emulate 68k apps via OS4. If OS4 had some worthwhile apps, then I could see why, but there just isn't any compelling reason to pay thousands for an outdated, feature-lacking OS to emulate 68k apps on outdated hardware. Mental illness maybe?

So, 68k obsession is hobby, PowerPC obsession mental illness. This logic is "perfect".


PowerPC applications I use: Odyssey, SMTube/MPlayer, AmiPDF, AbiWord, Gimp, PPaint (and other like emulators/interpreting engines etc.).
68k applications I use: TvPaint, IBrowse (for experiments only), MysticView (mixed with PPC libraries) and some small utils like FreeWheel and BarClock.

@haywiresystem

Quote:

3,000$ only gets you access to one core in amiga os. What a joke.

Sad to hear you are not enjoying your X1000.

@OlafS25

Quote:

Aros offers this... it depends where you run it

AROS68k... didn´t we speak about dead CPU architecture?

@ferrels

I think people that today use PPC with AmigaOS and MorphOS are long-term amigans who wanted faster hardware than the old 68k hardware that still offers a similar feeling and that is only possible with direct integration like with Petunia.

The MorphOS devs have already announced that they will use the ISA change for drastic improvements and thus changes at the OS including 64bit and SMP (about MP I cannot remember but would assume too). So such a modernized OS will break compatibility to existing MorphOS and 68k integration can only be done by using UAE.

The general problem I see when changing too much is that existing software cannot be just recompiled but has certainly partly be adapted/rewritten for the new platform. The best time would have been many years ago when there are still enough 3rd party developers who would have done that. Today two different MorphOSs as example would mean you need to develop for 2 different platforms to support both. There are not that many devs now left and from my experience becaus all do it as hobby most only support one platform. So I would predict that too drastic changes will lead to a fragmented MorphOS platform. But perhaps I am wrong .

There are different groups now, we have the vast majority of amigans who uses their old 68k amigas or emulation, then the "NG" user who use AmigaOS and MorphOS or AROS (i am a little beween with Aros 68k), then former Amiga-User and user that never heard of Amiga at all.

The last group today is of course the biggest, to reach this group a potentially modern OS is necessary and additionally you need new unique services and software.

For the other groups a modernized but still amigan OS running on modern hardware (like what people try with Aros) is better suited. PPC is only for the existing small group of hardcore fans and has no real future and potential to win new users. There are reasons why the MorphOS team will make a ISA change in future.

@pavlor

no it runs on my 2.7 Ghz system

and it outperforms everything AmigaOS runs on except X1000

@OlafS25

Quote:

PPC is only for the existing small group of hardcore fans and has no real future and potential to win new users.

I could say the same about 68k...

@OlafS25

Quote:

no it runs on my 2.7 Ghz system

Quote:

and it outperforms everything AmigaOS runs on except X1000

What is your "dnetc -benchmark" score?

@pavlor

I did not claim that 68k will become mainstream again, didn´t I?

i wrote that the vast majority in the community uses it on old hardware or emulation. Most do not own "NG"

@pavlor

what does this benchmark measure?

@OlafS25

OGRNG is good for raw integer performance, from my experience - it scales well with CPU speed. As AltiVec has clear advantage here, we can limit our comparison to non-AltiVec equipped PowerPC CPUs (440, G3 etc.).

Last edited by pavlor on 13-Jun-2015 at 09:15 AM.

@pavlor

I would test it if possible

any 68k version?

@OlafS25

Here


Simply type in CLI "dnetc -benchmark" and report OGRNG result.

@pavlor

the archive has a problem

@pavlor

Quote:

OGRNG is good for raw integer performance, from my experience - it scales well with CPU speed. As AltiVec has clear advantage here, we can limit our comparison to non-AltiVec equipped PowerPC CPUs (440, G3 etc.).

.

No, it is not. PA6-T at 1,8Ghz is slower than my previous G4 at 1,26 Ghz.

Futaura knows about it, we exchange abouth this and there is nothing he can do about it.

Here are the benchmarks.

OGR-NG with G4@1,26 Ghz:
OGR-NG: using core #1 (KOGE 3.1 Hybrid).
OGR-NG: Benchmark for core #1 (KOGE 3.1 Hybrid)
0.00:00:17.25 [26,837,018 nodes/sec]

OGR-NG with PA6T@1,8 Ghz:
OGR-NG: using core #1 (KOGE 3.1 Hybrid).
OGR-NG: Benchmark for core #1 (KOGE 3.1 Hybrid)
0.00:00:16.24 [22,526,735 nodes/sec]

RC5-72 with G4@1,26 Ghz:
RC5-72: using core #4 (KKS 7450).
RC5-72: Benchmark for core #4 (KKS 7450)
0.00:00:16.08 [13,142,653 keys/sec]

RC5-72 with PA6T@1,8 Ghz:
RC5-72: using core #4 (KKS 7450).
RC5-72: Benchmark for core #4 (KKS 7450)
0.00:00:16.80 [9,770,362 keys/sec]

As you can see, PA6T is way slower than the G4 which has a frequency slower
than the PA6T.

_________________
PowerMac G5 2,7Ghz - 2GB - Radeon 9650 - MorphOS 3.14
AmigaONE X1000, 2GB, Sapphire Radeon HD 7700
FPGA Replay + DB 68060 at 85Mhz