Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

[FAQ] PowerPC-Mac Benchmark link list/index

Status
The first post of this thread is a WikiPost and can be edited by anyone with the appropiate permissions. Your edits will be public.
C

Cox Orange

macrumors 68000
Original poster
Jan 1, 2010
1,814
241
Hi,

why this, can't I/people google? Every now and then, when we talk about benchmarks here or I want to know look up something, I have to search what I need, again.
Google? Well, let's say I only google barefeats benchmarks. For some reason it doesn't find what I was aiming for. Merely I get there by luck. Xlr8yourmac has deleted the nice index table for CPU-Upgrades for older models (only Mac Pros now), the old articles are there though. No way to easily find them.

This is why I decided to do this list here. If it doesn't help anyone else, it at least serves as a note for me.
From time to time I will have a look here and add the benchmarks, that others post in the OP. (it may take some time though - MODs, if it is possible you can give write-rights to everyone for my OP!!!)

@ ALL, please make recomendations, what I can do better in the OP, so that it gets better readable!

[R] = real life benchmarks (give seconds to complete or FPS)
[S] = synthetic benchmarks that just give points/numberscores (Geekbench)
[R/S] = mixed


[S] http://www.everymac.com/ultimate-mac-comparison-chart/?compare=all-macs
Geekbenchscores 32bit/64bit single/dual PPC/Intel

[R] https://forums.macrumors.com/threads/1862039/
Macs from this Forum, running Handbrake encoding.

[R] http://barefeats.com/fcp4.html
Final Cut Pro 4, 7 tests: PM G5, MP G4, PB G4, TiBook

[R] http://www.macworld.com/article/1047649/dualcorebenchmarks.html
early 2005 and late 2005 PowerMac G5 all models (despite Quad).

[R] http://barefeats.com/imacg5.html
Pmac G5/2.0 SE = G5/2.0GHz MP Power Mac with Radeon 9800 Pro SE
Pmac G4/1.42 MP = G4/1.42GHz MP Power Mac with Radeon 9700 Pro
Pbook G4/1.5 = G4/1.5GHz PowerBook 17" with Radeon 9700 mobility
iMac G5/1.8 = iMac G5/1.8GHz with GeForceFX 5200 (20" screen)
iMac G4/1.25 = iMac G4/1.25GHz with GeForceFX 5200 (20" screen)

[R] http://barefeats.com/g4up.html
Gigadesign/Sonnet 1.7-2GHz G4 vs MDD vs G5 2GHz

[R] http://barefeats.com/mini01.html
Mac Mini G4 1,42GHz and similarly clocked iMac, PowerMac and Cube-(with Upgrade)

[R] http://barefeats.com/imcd2.html
Intel-iMac vs PM G5, iMac G5 in Gaming: Unreal, Doom3, WoW


Aftermarket Upgrades:
list of CPU-Upgrades for G4s http://beta.ivc.no/wiki/index.php/PowerMac_G4_Upgrades#Upgrade_boards
G4s http://lowendmac.com/ppc/power-mac-g4-upgrade-guide.html
MDD G4s http://lowendmac.com/2007/3-cpu-upgrades-for-mirrored-drive-doors-g4-power-macs/
G4 and G3 ZIF http://lowendmac.com/ppc/g4zif.shtml
Powerbook Wallstreet/Lombard http://lowendmac.com/ibook/010316.html
G3s http://lowendmac.com/ppc/g3card.shtml

[R] 7447/7448 (7455 stock) http://barefeats.com/g4up2.html
7448 1.8 MP / 7448 2.0 SP / 7447A 2.0 SP in QS
PM G5 2.0 MP b / a = Power Mac G5/2.0GHz "June 2003" with b) Radeon X800 XT / a) Radeon 9800 Pro SE
iMac(Intel) CD 2.0
iMac G5 2.1 "iSight" /2.0GHz "ALS"
stock MDD-FW800 (7455) 1.42 MP
QS 1.0 MP, 800 MP

[R/S] http://www.xlr8yourmac.com/G4CARDS/powerlogix_dual_g4_7447_7457/index2.html
7447, 7457, stock QS

[R] http://barefeats.com/cubeup.html
Cube 500MHz vs Cube-Upgrade (1.3-1.7GHz, 7447, 7457) vs Mac Mini.


(7300-9600 Mac systems)
http://www.xlr8yourmac.com/G4CARDS/


EDIT: until I find my GPU benchmark collection list with benchmarks from barefeats and other xlr8yourmac tests I leave this here http://www.xlr8yourmac.com/Graphics/Radeon_9700PRO_OEM/index.html
 
Last edited:
Cox Orange said:
@ ALL, please make recomendations, what I can do better in the OP, so that it gets better readable!
Click to expand...

I think geekbench is a terrible benchark for CPU. Here is a plot for a single type of computer (a 12" 1.5GHz powerbook):

2a8j1vq.png


That really pins down a number doesn't it? :rolleyes: In my opinion nbench is a far superior benchmark for CPU comparison. It's multi-platform and actually does some statistics internally as it runs. For example, I can run the benchmark on a clean boot and get the same numbers as I get while running it while I watch a movie at the same time. It doesn't have a fancy front end though. Here is the info and a whole slew of results to compare against:

http://www.tux.org/~mayer/linux/bmark.html

http://www.tux.org/~mayer/linux/results2.html
 
flyrod said:
I think geekbench is a terrible benchark for CPU. Here is a plot for a single type of computer (a 12" 1.5GHz powerbook):

Image
Click to expand...

Unfortunately, mine only scored 562 due to having a dead battery, which causes the CPU to run at a lower speed in order to keep power consumption down:

attachment.php
 
Last edited:
flyrod said:
I think geekbench is a terrible benchark for CPU. Here is a plot for a single type of computer (a 12" 1.5GHz powerbook):

Image

That really pins down a number doesn't it? :rolleyes: In my opinion nbench is a far superior benchmark for CPU comparison. It's multi-platform and actually does some statistics internally as it runs. For example, I can run the benchmark on a clean boot and get the same numbers as I get while running it while I watch a movie at the same time. It doesn't have a fancy front end though. Here is the info and a whole slew of results to compare against:

http://www.tux.org/~mayer/linux/bmark.html

http://www.tux.org/~mayer/linux/results2.html
Click to expand...

I don't suppose there's a readymade OS X executable knocking about is there ? :D
 
Dronecatcher said:
I don't suppose there's a readymade OS X executable knocking about is there ? :D
Click to expand...

I compiled this a few years ago for G4 (but no altivec) using Xcode. Different compilers or compiler options would change things slightly. For reference, here's my mini results:
Code: 
TEST                : Iterations/sec.  : Old Index   : New Index
--------------------:------------------:-------------:------------
NUMERIC SORT        :          949.37  :      24.35  :       8.00
STRING SORT         :           241.6  :     107.95  :      16.71
BITFIELD            :      2.5323e+08  :      43.44  :       9.07
FP EMULATION        :          162.44  :      77.94  :      17.99
FOURIER             :           14982  :      17.04  :       9.57
ASSIGNMENT          :          28.777  :     109.50  :      28.40
IDEA                :          4179.9  :      63.93  :      18.98
HUFFMAN             :          1559.7  :      43.25  :      13.81
NEURAL NET          :          16.891  :      27.13  :      11.41
LU DECOMPOSITION    :             700  :      36.26  :      26.19
==========================ORIGINAL BYTEMARK RESULTS==========================
INTEGER INDEX       : 59.673
FLOATING-POINT INDEX: 25.593
Baseline (MSDOS*)   : Pentium* 90, 256 KB L2-cache, Watcom* compiler 10.0
==============================LINUX DATA BELOW===============================
CPU                 : 1.75GHz G4 Mac Mini
L2 Cache            : 512KB
OS                  : Darwin 9.8.0
C compiler          : gcc version 4.0.1 (Apple Computer, Inc. build 5370)
libc                : 
MEMORY INDEX        : 16.269
INTEGER INDEX       : 13.934
FLOATING-POINT INDEX: 14.195

I am curious if gcc 4.9 would be any better now than the old Xcode.
 

Attachments

  • nbench.zip
    40.2 KB · Views: 189
Dronecatcher said:
On my G5 & Powerbook it stops at the Neural Net test with :
CPU:NNET--error in opening file!
Click to expand...

The neural net test uses the file NNET.DAT to read in the network. I think I included this file in the zip. From a terminal, run nbench from the same directory as NNET.DAT so it can find the file...
 
flyrod said:
The neural net test uses the file NNET.DAT to read in the network. I think I included this file in the zip. From a terminal, run nbench from the same directory as NNET.DAT so it can find the file...
Click to expand...

To be on the safe side, I copied both nbench and NNET.DAT into the Utilities folder where Terminal resides and ran it from there but still get the same error.
 
OK, who crossed out ALL links and categories? and more important, WHY!!!???!!!
 
I mean, perhaps it is even legitimate to cross out everything, because it was stupid from me to start such a thread. The stupid one usually doesn't know he is stupid himself.
But if others see it as incorrect to cross out everything, as well, it is like everything in life: just because there are some people that don't behave, others have to suffer under restrictions. -> someone can destroy the text every now and then in the future, again, so one might have to remove the wiki-edit feature of this thread and people will have to post in the thread and I will put the info in the 1st post myself.

Also the person let "real life test" intact, but failed to realize that some of the benchmarks listed are done with real life scenarios.
 
Cox Orange said:
I mean, perhaps it is even legitimate to cross out everything, because it was stupid from me to start such a thread. The stupid one usually doesn't know he is stupid himself.
But if others see it as incorrect to cross out everything, as well, it is like everything in life: just because there are some people that don't behave, others have to suffer under restrictions. -> someone can destroy the text every now and then in the future, again, so one might have to remove the wiki-edit feature of this thread and people will have to post in the thread and I will put the info in the 1st post myself.

Also the person let "real life test" intact, but failed to realize that some of the benchmarks listed are done with real life scenarios.
Click to expand...

It is childish for someone to do it - but I'm sure there'll be a record of who altered it. On the plus side, you can still read the info and it's easy enough to take the lines out - try not to take it personally - it's just the actions of some coward who daren't even argue his case if he has a disagreement.
 
Heh, who could have thought about that? :) hm, and I was like making a conspiracy theory... :rolleyes:
 
After all of these years finally some benchmarks.
 

Attachments

  • OpenMark result.jpg
    OpenMark result.jpg
    41.2 KB · Views: 83
  • cinebenchg5.jpg
    cinebenchg5.jpg
    479.4 KB · Views: 84
  • geekbenchg532bit.jpg
    geekbenchg532bit.jpg
    465.6 KB · Views: 87
  • xbenchg5.jpg
    xbenchg5.jpg
    409.7 KB · Views: 75
  • geekbenchg564bit.jpg
    geekbenchg564bit.jpg
    464.1 KB · Views: 87
Thanks for posting your results. Here is some more info:


Here is some CPU testing stuff:

sourceforge.net

iBench

Download iBench for free. The Cocoa Benchmark. iBench is a benchmark for Mac OS X constituted by a suite of tests of real-world usage that allows you to check and compare Macintosh computers performance on integer and floating point workloads. iBench is compatible with both Intel and PowerPC Macs.
sourceforge.net sourceforge.net
(could someone compile this for OS X 10.4? I get errors trying to do it...)

The new home of nbench:


Newer Technology still has some stuff on their web site:




This thread has some gcc testing I did on a few G4s:


Here is some hard drive testing:

forums.macrumors.com

Post Your SSD/CF SATA/PATA PowerPC Benchmark Results

It would be interesting to compile a collection of real world PowerPC benchmark results with different configurations of solid state drives from factory sealed 2.5" PATA SSDs, mSATA and the various adapter caddies and CF solutions. There is a lot of information floating around, but nothing that...
forums.macrumors.com forums.macrumors.com

Here is some video card testing:

forums.macrumors.com

Some PPC OpenMark benchmarks

I got bored and ran GeoFX Openmark 1.6 OpenGL benchmark on some of my PPC macs, the results were interesting. Now including measurements for other PowerPC macs from forum members! Now a WIKI! Thanks to all who helped create this list! ATI Rage Mobility 128 ---------139 (G3 Pismo Powerbook...
forums.macrumors.com forums.macrumors.com
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back