@monomango

Hi!
I overclocked my old A1SE to 667 MHz just after buying it and used it for about 7 years without big problems I used an ATA133 Sil card and onboard ethernet.
I only modified the moltiplcator without touching the core voltage.

@Thematic

if you still need here Peg2 G4

@328gts


uhm...
still slow the dnetc..

Can you plz try to run that way:

dntec -bench

so all dntec cores was tested.

And plz add also lame test.

Thank you.

Hello all,
I bought that:

https://it.aliexpress.com/item/Nygacn- ... Computer/32668351496.html

on ebay and unfortunately seems to not works on my Pegasos2 AmigaOs4.1FEUpd1 setup...instead it works perfectly under MOS...any hope to have it working on AmigaOS4.1 someway?

@328gts

I updated the main topic on AW, but still waiting for your benchmarks on celan start to update completely.
The main "propbem" seems with ORG-NG/RC5-72 that was really slow...
Dunno if was normal for PA6T or not...however on my Peg2 was faster...

@AmigaBlitter

but wich type of video?

For simple videos you can use Hollywood :)

@328gts

better pressing "ALT" while loading WB so no WBStartup loaded :)

all, works ok here, I opened the program, played here and there, changed screenmode nad paint size, no problems at all...

@328gts

thank you for results.
Have you run the tests at clean start(i.e. without anything other loaded?)

We are really lacking x5000 and x1000 benchs(also Vampire if someone can do).


Plz guys help.

updated, plz update also that topic.

On my Pegasos2 G4@1131MHz / Radeon9000pro i got:

GPmark: width 640, height 480, flags 0x80000000
Blitting Test: 234.2
Plasma: 183.1
Rotozoomer: 214.5
Rotozoomer Near: 221.4
Rotozoomer Far: 214.2
Radial Blur: 21.4
3D Bunny: 11.5

@LiveForIt

exactly

since I opened tehe topic(as you probably know):

http://amigaworld.net/modules/newbb/v ... t=0&viewmode=flat&order=0

And want to make tests much exactly possible, I found that GPMark with about same date on AmiNET, OS4Depot, and AROS Archives so thinked that they was about same "compile" and good to do tests...
However now that thread make me doubts on that...
Can you maybe check on the above archives if they can be compared effectively or better to remove the benchmark from my thread plz if that wasnt so complex?

Thank you for any statement/help on that.

@Capehill

mmm...
dunno how a gpmark benchmarks have to be compiled...

In theory it have to be much standard as possible to compare results on every system compiled right?

I man on my benchamerk test if we use gpmark to made comparisions it have to be compiled "same" way on every system....

How to do that?

Do you think your compile was much standard than the other? or not?

Can you check someway the os3.x/mos/aros ports to see compile differences?

Another update with non altivec results for better and clear comparison with x5000 and WinUAEppc for exaple.

@all

I've updated the topic trying to put a clear "table" of posted results on originale thread.
If someone can do some more test and also someX1000/x5000/Viper/WinUAE/MOS can do some more test is welcome

PS @328gts

plz copy/paste the new thread op on that thred op if you can, thank you.

@328gts

Updated topic with dntec benchmark:

Download exes:

http://http.distributed.net/pub/dcti/ ... ent/dnetc-amigaos-ppc.lha ---> dntec AmigaOS4

http://http.distributed.net/pub/dcti/ ... ent/dnetc-morphos-ppc.lha ---> dntec MorphOS

http://http.distributed.net/pub/dcti/ ... ent/dnetc-amigaos-68k.lha ---> dntec AmigaOS68k

http://http.distributed.net/pub/dcti/ ... dnetc-amigaos-ppc-wos.lha ---> dntec WarpOS

my AmigaOS4.1 results:

netc v2.9112-521-CTR-16020313 for AmigaOS (OS 4.1, PowerPC).
Please provide the *entire* version descriptor when submitting bug reports.
The distributed.net bug report pages are at http://bugs.distributed.net/

[Jan 01 18:24:28 UTC] Automatic processor type detection found
a PowerPC 7445/7455 (G4) processor.
[Jan 01 18:24:28 UTC] OGR-NG: using core #1 (KOGE 3.1 Hybrid).
[Jan 01 18:24:49 UTC] OGR-NG: Benchmark for core #1 (KOGE 3.1 Hybrid)
0.00:00:17.39 [24,052,847 nodes/sec]
[Jan 01 18:24:49 UTC] RC5-72: using core #4 (KKS 7450).
[Jan 01 18:25:08 UTC] RC5-72: Benchmark for core #4 (KKS 7450)
0.00:00:16.70 [11,832,753 keys/sec]
[Jan 01 18:25:08 UTC] Compare and share your rates in the speeds database at
http://www.distributed.net/speed/
(benchmark rates are for a single processor core)

to perform test use:

dntec -benchmark

@samo79

but if you put the beta in the os4depot version drawer there's no output and its also faster...
did you tryed?

ok upload finished...

https://app.box.com/s/uj6xruq43lqtbdu6801j56c73le9hn4n

in the folder was present 3 video files.


mp4 file have thye Home directory problem
flv video ifle have the ahi crash with os4depot version
wmv file dont have the ahi problem with os4depot version and was the file I used to benchmark

Hope that helps

EDIT: ok, mp4 file seems corrupt...no way to play it with any mplayer version of ffplay...

EDIT2: I uploaded QauntumRise.vob converted from the wmv file using ffmpeg gui with latest ffmpeg from os4depot, I wnat only to point that the vob file was played by afxgroup about same speed of other mplayer ports(related to cpu usage looking to cpu docky), but benchmarking it I got 93sec instead of 30/32 of mickJT version, liveforIT version instead geve me "[mpeg2video @ 0x6063d460]get_buffer() failed (stride changed)
Error while decoding frame!" but plays it correctly if playing it normally.

Also have a file where video was corrupt(like the lines wasnt ended correctly so the video was all corrupted like if tehe previous line was ending on the next line for few pixels and so on or something like that) afxgroup mplayer instead play it correctly(afx group mplayer play it correctly)...was a flv video downloaded from YouTube(too big to upload...about 130MB, but if needed I'll try)

EDIT3: ok I splitted the file was the "Why use amiga..."