Poster | Thread |
jPV
| |
Re: Blender benchmark time! Posted on 26-Jan-2010 12:42:40
| | [ #41 ] |
|
|
|
Cult Member |
Joined: 11-Apr-2005 Posts: 822
From: .fi | | |
|
| With one quick try on Mac mini G4/1.5GHz and MorphOS:
Blender 2.46: 06:22:90 Blender 2.5 alpha 0: 06:58:93
_________________ - The wiki based MorphOS Library - Your starting point for MorphOS - Software made by jPV^RNO |
|
Status: Offline |
|
|
ddni
| |
Re: Blender benchmark time! Posted on 26-Jan-2010 13:08:25
| | [ #42 ] |
|
|
|
Cult Member |
Joined: 11-Jan-2007 Posts: 818
From: Northern Ireland | | |
|
| Has anyone submitted their Amiga times to the site yet? I cant see any Amiga results.
_________________ AmigaOne X1000 |
|
Status: Offline |
|
|
Tuxedo
| |
Re: Blender benchmark time! Posted on 26-Jan-2010 13:23:38
| | [ #43 ] |
|
|
|
Elite Member |
Joined: 28-Nov-2003 Posts: 2346
From: Perugia, ITALY | | |
|
| @jPV
Nice results! just for curiosity...tested same Blender versions on MacOSX?
_________________ Simone"Tuxedo"Monsignori, Perugia, ITALY. |
|
Status: Offline |
|
|
Karlos
| |
Re: Blender benchmark time! Posted on 26-Jan-2010 13:27:53
| | [ #44 ] |
|
|
|
Elite Member |
Joined: 24-Aug-2003 Posts: 4415
From: As-sassin-aaate! As-sassin-aaate! Ooh! We forgot the ammunition! | | |
|
| |
Status: Offline |
|
|
KimmoK
| |
Re: Blender benchmark time! Posted on 26-Jan-2010 13:28:07
| | [ #45 ] |
|
|
|
Elite Member |
Joined: 14-Mar-2003 Posts: 5211
From: Ylikiiminki, Finland | | |
|
| @pavlor
"It runs in WMWare..."
oops _________________ - KimmoK // For freedom, for honor, for AMIGA // // Thing that I should find more time for: CC64 - 64bit Community Computer? |
|
Status: Offline |
|
|
broadblues
| |
Re: Blender benchmark time! Posted on 26-Jan-2010 13:33:16
| | [ #46 ] |
|
|
|
Amiga Developer Team |
Joined: 20-Jul-2004 Posts: 4446
From: Portsmouth England | | |
|
| @ddni
Quote:
Has anyone submitted their Amiga times to the site yet? I cant see any Amiga results
|
I have, and afew others, but I think they have to be manually checked before inclusion, so it might take the site maintainer a while if he's busy etc.
[edit]reversed m and n in namually [/edit]Last edited by broadblues on 26-Jan-2010 at 02:12 PM.
_________________ BroadBlues On Blues BroadBlues On Amiga Walker Broad |
|
Status: Offline |
|
|
klx300r
| |
Re: Blender benchmark time! Posted on 26-Jan-2010 13:36:04
| | [ #47 ] |
|
|
|
Elite Member |
Joined: 4-Mar-2008 Posts: 3839
From: Toronto, Canada | | |
|
| @ddni
I have too!..it's cool to see Amiga on the list ....sure we're not the quickest but we can still play the game well _________________ ____________________________ c64-2sids, A1000, A1200T-060@50(finally working!),A4000-CSMKIII ! My Master Miggies- Amiga 1000 & AmigaOne X1000 ! mancave-ramblings X1000 I BELIEVE |
|
Status: Offline |
|
|
KimmoK
| |
Re: Blender benchmark time! Posted on 26-Jan-2010 13:41:25
| | [ #48 ] |
|
|
|
Elite Member |
Joined: 14-Mar-2003 Posts: 5211
From: Ylikiiminki, Finland | | |
|
| @jPV
"Blender 2.46: 06:22:90"
So thats around Intel® Pentium® 4 2666MHz and AMD Sempron™ 3600+ 2000MHz speed. Not too bad.
I wonder how close 2Ghz PA6T would be to 03:42.93 of AMD Athlon™64 X2 4600+ 2400MHz...
Perhaps even faster.
(Dualcore 1,5Ghz G4 would be 3:11.5 if it would scale 1:1) Last edited by KimmoK on 26-Jan-2010 at 01:43 PM. Last edited by KimmoK on 26-Jan-2010 at 01:41 PM.
_________________ - KimmoK // For freedom, for honor, for AMIGA // // Thing that I should find more time for: CC64 - 64bit Community Computer? |
|
Status: Offline |
|
|
bitman
| |
Re: Blender benchmark time! Posted on 26-Jan-2010 14:53:49
| | [ #49 ] |
|
|
|
Cult Member |
Joined: 25-Mar-2008 Posts: 705
From: Fredericia, Denmark | | |
|
| |
Status: Offline |
|
|
broadblues
| |
Re: Blender benchmark time! Posted on 26-Jan-2010 14:57:30
| | [ #50 ] |
|
|
|
Amiga Developer Team |
Joined: 20-Jul-2004 Posts: 4446
From: Portsmouth England | | |
|
| @bitman
they aren't there yet. Give the guy a chance he's just been hit with a load of entries from an operating system, he may not even have heard of, at least not in it's modern incarnation.
_________________ BroadBlues On Blues BroadBlues On Amiga Walker Broad |
|
Status: Offline |
|
|
$adddam
| |
Re: Blender benchmark time! Posted on 26-Jan-2010 15:21:54
| | [ #51 ] |
|
|
|
Regular Member |
Joined: 24-May-2006 Posts: 194
From: magyarorszag /=hungary/ | | |
|
| 10:05:75
Peg2/G4, MorphOS 2.4
i dont understand the results btw. in my leauge there are: CoreDuo 2,9Ghz - 10:00,79 Athlon 64 3Ghz - 10:04,00 P4 2,8Ghz - 10:05,23 UltraSpacII 450Mhz!- 10:05,78 ! even a 733MHz p3 achieved similar results. how can it be possible? according to these results even a subGhz athlon thunderbird can beat a 1.4Ghz G4 (and both can beat coreduo, athlon64)... doesnt make sense...
Last edited by $adddam on 26-Jan-2010 at 03:23 PM.
|
|
Status: Offline |
|
|
klx300r
| |
Re: Blender benchmark time! Posted on 26-Jan-2010 15:29:16
| | [ #52 ] |
|
|
|
Elite Member |
Joined: 4-Mar-2008 Posts: 3839
From: Toronto, Canada | | |
|
| @broadblues
@ $addam
of course you're "assuming" people are honest with their times _________________ ____________________________ c64-2sids, A1000, A1200T-060@50(finally working!),A4000-CSMKIII ! My Master Miggies- Amiga 1000 & AmigaOne X1000 ! mancave-ramblings X1000 I BELIEVE |
|
Status: Offline |
|
|
itix
| |
Re: Blender benchmark time! Posted on 26-Jan-2010 15:32:19
| | [ #53 ] |
|
|
|
Elite Member |
Joined: 22-Dec-2004 Posts: 3398
From: Freedom world | | |
|
| I think results depend on Blender version, number of CPUs and RAM. _________________ Amiga Developer Amiga 500, Efika, Mac Mini and PowerBook |
|
Status: Offline |
|
|
Karlos
| |
Re: Blender benchmark time! Posted on 26-Jan-2010 15:42:00
| | [ #54 ] |
|
|
|
Elite Member |
Joined: 24-Aug-2003 Posts: 4415
From: As-sassin-aaate! As-sassin-aaate! Ooh! We forgot the ammunition! | | |
|
| @itix
Blender doesn't automatically scale to more CPUs in versions before 2.6, you have to tell it how many threads to use.
I use 8 threads on my Core 2 Quad having observed that a single thread did not use the full capacity of one core and actually observed a ~2.5 second speed up (about 10%) compared to using 4 threads. That resulted in a pretty respectable timing of 25.34 seconds :) Last edited by Karlos on 26-Jan-2010 at 03:43 PM.
_________________ Doing stupid things for fun... |
|
Status: Offline |
|
|
Fernecho
| |
Re: Blender benchmark time! Posted on 26-Jan-2010 15:51:25
| | [ #55 ] |
|
|
|
Member |
Joined: 4-Jul-2009 Posts: 80
From: TRELLE ( Ourense - Spain ) | | |
|
| @ddni
I have sent my results yesterday , the results take 24 hours to leave, my results come out tonight.
_________________ AmigaONE X1000, Radeon HD 7950 Amiga 1200 / Blizzard 1260 / Mediator 1200 / Voodoo 3000 / 64 Mb / AmigaOS 3.9 |
|
Status: Offline |
|
|
K-L
| |
Re: Blender benchmark time! Posted on 26-Jan-2010 17:54:50
| | [ #56 ] |
|
|
|
Super Member |
Joined: 3-Mar-2006 Posts: 1413
From: Oullins, France | | |
|
| @Thread
AmigaOne G4 1Ghz (7447 from Acube) : 00:10:05.82 Opteron180 (with Mandrva 2010) 2,72 Ghz (1 CPU used) : 00:02:11.11
Is Altivec used with the AmigaOS version of Blender ?
Last edited by K-L on 26-Jan-2010 at 06:24 PM.
_________________ PowerMac G5 2,7Ghz - 2GB - Radeon 9650 - MorphOS 3.14 AmigaONE X1000, 2GB, Sapphire Radeon HD 7700 FPGA Replay + DB 68060 at 85Mhz |
|
Status: Offline |
|
|
bernd_afa
| |
Re: Blender benchmark time! Posted on 26-Jan-2010 18:23:17
| | [ #57 ] |
|
|
|
Cult Member |
Joined: 14-Apr-2006 Posts: 829
From: Unknown | | |
|
| @KimmoK > wonder how close 2Ghz PA6T would be to 03:42.93 of AMD Athlon™64 X2 4600+ >2400MHz...
The pa6T is also a slow PPC CPU and i think not reach the performance /MHZ as the IBM G3 or freeescale G4 can reach.
See here the benchmark that compare pa6t with a AMD low power mobility CPU. pa6 is from 30% and more slower as a actual AMD embedded CPU that have much fewer Cache as the PA6.
http://www.coremark.org/benchmark/index.php?pg=benchmark
to show the value of pa6 you need check it on on left side.Note: pa6 is 2 threads so you should compare the values with 2 thread values of AMD CPU
but to say more, more benchmarks of Pa6 are usefull.Have somebody see more benchmarks of Pa6 ?
Last edited by bernd_afa on 26-Jan-2010 at 06:29 PM. Last edited by bernd_afa on 26-Jan-2010 at 06:25 PM. Last edited by bernd_afa on 26-Jan-2010 at 06:24 PM.
|
|
Status: Offline |
|
|
bernd_afa
| |
Re: Blender benchmark time! Posted on 26-Jan-2010 18:40:33
| | [ #58 ] |
|
|
|
Cult Member |
Joined: 14-Apr-2006 Posts: 829
From: Unknown | | |
|
| @$adddam
>according to these results even a subGhz athlon thunderbird can beat a 1.4Ghz G4 >(and both can beat coreduo, athlon64)... doesnt make sense...
You should look at the thread values.when see there is a 8 thread machine at 3,4 GHZ that need to 9 minutes.so it seem much threads on blender slow things down.
The P3 733 MHZ results are usefull, because they show that with that settings it can run faster
GCC -O3, SSE, fast-math
maybe broadblues compile blender with -O3 and fast-math new |
|
Status: Offline |
|
|
broadblues
| |
Re: Blender benchmark time! Posted on 26-Jan-2010 19:00:41
| | [ #59 ] |
|
|
|
Amiga Developer Team |
Joined: 20-Jul-2004 Posts: 4446
From: Portsmouth England | | |
|
| @K-L
Quote:
Is Altivec used with the AmigaOS version of Blender ?
|
No. There's no altivec code in the main blender source tree, so the only benefits would be in compiler optimisation, I don't know how much difference that would make.
The built in ffmpeg can be built with altivec, (and was initially when I was developing on my old AOne), but would make little difference to rendering this benchmark.
altivec and 440ep extentions are used indirectly via newlib ofcourse.
I may do a altivec optimised build for the next release, but I have no way of verifying that it's built correctly. I'm undecided at this point in time as to it's usefullness compared to effort requred (blender takes agaes to build from scratch)
Last edited by broadblues on 26-Jan-2010 at 07:01 PM.
_________________ BroadBlues On Blues BroadBlues On Amiga Walker Broad |
|
Status: Offline |
|
|
Karlos
| |
Re: Blender benchmark time! Posted on 26-Jan-2010 19:09:13
| | [ #60 ] |
|
|
|
Elite Member |
Joined: 24-Aug-2003 Posts: 4415
From: As-sassin-aaate! As-sassin-aaate! Ooh! We forgot the ammunition! | | |
|
| @broadblues
There are SSE2 optimisations available. Apple produced some fairly detailed documentation on how to migrate altivec code to SSE, I wonder if it could be applied in reverse as a hint as to what to optimise and how? _________________ Doing stupid things for fun... |
|
Status: Offline |
|
|