Power Mac G5 Quad Benchmarks

[Pie]

Yeah. Looks like once we get quad 3.0Ghz machines, I'm upgrading my current Power Mac. That I promised myself.

However, with the intels on their way, my plans for upgrading have been thrown a monkey wrench. What to do? What to do.

What swung it for me was the fact I didn't want to wait for compliant dual binaries to install on the Intels. And those rumblings of on-chip DRM.

 

[EdwinSneller]

Here is my Dual 2.7 Cinebench score:

CINEBENCH 2003 v1
****************************************************

Tester : sneller

Processor : PowerMac G5 Dual
MHz : 2.7
Number of CPUs : 2
Operating System : 10.4.2

Graphics Card : ATI 9650 256MB
Resolution : 1680x1050
Color Depth : millions

****************************************************

Rendering (Single CPU): 385 CB-CPU
Rendering (Multiple CPU): 663 CB-CPU

Multiprocessor Speedup: 1.72

Shading (CINEMA 4D) : 359 CB-GFX
Shading (OpenGL Software Lighting) : 1028 CB-GFX
Shading (OpenGL Hardware Lighting) : 1736 CB-GFX

OpenGL Speedup: 4.83

****************************************************

 

[rickvanr]

Ok, how does Cinebench work? Don't take this seriously, I have no idea how Cinebench works.

CINEBENCH 2003 v1
****************************************************

Tester : Rick

Processor : PowerMac G5
MHz : 2.7 (2)
Number of CPUs : 2
Operating System : 10.4.2

Graphics Card : 6800 Ultra
Resolution : 1920 x 1200
Color Depth : Millions

****************************************************

Rendering (Single CPU): 294 CB-CPU
Rendering (Multiple CPU): 537 CB-CPU

Multiprocessor Speedup: 1.82

Shading (CINEMA 4D) : 341 CB-GFX
Shading (OpenGL Software Lighting) : 904 CB-GFX
Shading (OpenGL Hardware Lighting) : 1490 CB-GFX

OpenGL Speedup: 4.37

****************************************************

 

[JoeBeCrazy]

how much ram do you have? i ahve a new 15'' and 512 ddr2 and it gets a lot lower

Yea i have the New Powerbook 15" 1.67" with 2GB ram... i get way lower than that.

 

[steve_hill4]

Yeah. Looks like once we get quad 3.0Ghz machines, I'm upgrading my current Power Mac. That I promised myself.

However, with the intels on their way, my plans for upgrading have been thrown a monkey wrench. What to do? What to do.

A lot of professional Mac users, (music business, video editing), have said the best machine to get is the quad 2.5 as it should last any user for a good 3 years. A quad 3.0 would do the same job, but get released just before the intel switch, (assuming two more upgrades before 2.5>2.7>3.0), so you would be looking at buying late next year, early 2007 and replace with a MacIntel in 2010.

I don't think there are currently sufficient advantages with the intel over G5 ranges, G4 yes, G5 no. Not sure what intel's roadmap holds for the period of 2007-2010, but can't see too much that would theoretically beat a quad 3.0.

 

[EdwinSneller]

FYI: it took Photoshop CS2 40 seconds to spin the horse on my Dual 2.7 GHz.

 

[melgross]

I know that the whole Xbench thing surely has to be messed up, otherwise it's basically close to 50% increase in speed over a Dual 2 (4GHz). Now linearly only a Dual 3 (6GHz) would be 50% faster than the Dual 2...so surely Quad 2.5s (10GHz) would be more than 50% faster.

I don't know. If you look at Macworlds tests they show the same thing.

My own PS tests on numerous machines also show that.

One machine I upgraded recently is a Digital Audio 733MHz G4, 256KB L2, 1MB L3.

Upgraded to dual 1.8GHz 7447a's.

Some tests

45MB photo

733MHz
Sharpen 100% R1, T2 = 6 sec
Resample 2x = 5
Gauss Blur 10 pix = 13
" 50 " = 15
Rotate 90' clock = 5
" 45' " = 15
" 5' " = 11


dual 1.8 GHz

sharpen = 4 sec
resample = 3
Gauss = 6
" = 7
Rotate = 3
" = 7
" = 5

Yes, it's the same bus. But it isn't RAM limited it's at 100% efficiency. 1.5GB RAM. ATI 9800 card.

The surprising thing is that my dual 2GHz is only about 35% faster than the dual, but I don't seem to have the numbers.

What is frustrating is that Macworld's numbers give the dual 2.7's a Speedmark score of 248, but the Mini with it's 1.25 G4 and slow memory bus is now 100!

so a dual 2.7 is only 2.5 times faster than a low end Mini.

Time for reflection.

 

[risc]

...I DID USE THE LATEST VERSION OF XBENCH (1.2)...

Pie why did you reply to me with this? My you can't get over 100 comments are about a guy with a 15" PowerBook. I like everyone here doesn't doubt that a quad will beat a dual 2.0, maybe you should of read my post before replying?

For comparison, my PowerBook (15" Alu) 1.25 G4 got a 102.80

Brian

 

[magor]

CINEBENCH 2003 v1
****************************************************

Tester :

Processor : G5
MHz : 2.7
Number of CPUs : 2
Operating System : 10.4.3

Graphics Card : X800
Resolution : <fill this out>
Color Depth : <fill this out>

****************************************************

Rendering (Single CPU): 378 CB-CPU
Rendering (Multiple CPU): 643 CB-CPU

Multiprocessor Speedup: 1.7

Shading (CINEMA 4D) : 354 CB-GFX
Shading (OpenGL Software Lighting) : 956 CB-GFX
Shading (OpenGL Hardware Lighting) : 1995 CB-GFX

OpenGL Speedup: 5.64

****************************************************

 

[Pie]

Pie why did you reply to me with this, my you can't get over 100 comments are about a guy with a 15" PowerBook. I like everyone here doesn't doubt that a quad will beat a dual 2.0, maybe you should of read my post before replying?

You're absolutely right.
I need to open my eyes and spend less time speed reading (that's an oxymoron isn't it?)

be cool,
Pie

 

[MacEyeDoc]

Cinebench on a 1.25GHz PB with 1.25 GB RAM

CINEBENCH 2003 v1
****************************************************

Tester : omeyemo

Processor : G4 PowerBook
MHz : 1.25 GHz
Number of CPUs : 1
Operating System : 10.4.2

Graphics Card : ATI Mob Rad 9600
Resolution : 1280 x 854
Color Depth : Millions

****************************************************

Rendering (Single CPU): 113 CB-CPU
Rendering (Multiple CPU): --- CB-CPU


Shading (CINEMA 4D) : 139 CB-GFX
Shading (OpenGL Software Lighting) : 388 CB-GFX
Shading (OpenGL Hardware Lighting) : 783 CB-GFX

OpenGL Speedup: 5.63

****************************************************

 

[gmnygts]

Now we need to get someone with a dual 2.7 to run the same thing - anyone?

D

If some people are wondering about their 1st gen PMG5
my old machine:

CINEBENCH 2003 v1
****************************************************

Tester : gmnygts

Processor : dual 2.0 rev. A
MHz : 2.0 GHZ
Number of CPUs : 2
Operating System : 10.4.3

Graphics Card : ATI 850xt
Resolution : <fill this out>
Color Depth : <fill this out>

****************************************************

Rendering (Single CPU): 286 CB-CPU
Rendering (Multiple CPU): 514 CB-CPU

Multiprocessor Speedup: 1.79x

Shading (CINEMA 4D) : 276 CB-GFX
Shading (OpenGL Software Lighting) : 820 CB-GFX
Shading (OpenGL Hardware Lighting) : 1436 CB-GFX

OpenGL Speedup: 5.20x

****************************************************
using Pie's numbers to calculate times(x) faster, using my machine as baseline;
Rendering (Single CPU): 1.255 faster
Rendering (Multiple CPU): 1.976 faster

Shading (CINEMA 4D) : 1.278 faster
Shading (OpenGL Software Lighting) : 1.281 faster
Shading (OpenGL Hardware Lighting) : 1.302 faster

glen

 

[plinden]

Just to throw in a PC cinebench test, on a machine running lots of other stuff and not restarted:
CINEBENCH 2003 v1
****************************************************

Tester : plinden

Processor : Thinkpad T41p
MHz : 1600 (Banias Core)
Number of CPUs : 1
Operating System : Windows XP Professional SP1

Graphics Card : ATI Mobility Fire GL 9000
Resolution : 1280x1028
Color Depth : 32-bit

****************************************************

Rendering (Single CPU): 188 CB-CPU
Rendering (Multiple CPU): --- CB-CPU


Shading (CINEMA 4D) : 187 CB-GFX
Shading (OpenGL Software Lighting) : 810 CB-GFX
Shading (OpenGL Hardware Lighting) : 1429 CB-GFX

OpenGL Speedup: 7.63

****************************************************

 

[MacEyeDoc]

My PB is lame

****************************************************
using Pie's numbers to calculate times(x) faster, using my machine as baseline;
Rendering (Single CPU): 1.255 faster
Rendering (Multiple CPU): 1.976 faster

Shading (CINEMA 4D) : 1.278 faster
Shading (OpenGL Software Lighting) : 1.281 faster
Shading (OpenGL Hardware Lighting) : 1.302 faster

glen[/QUOTE]


And the Quad G5 is 8.99 times (you could call it 900% faster) than my PowerBook. Wow! Gotta get me one of those!

 

[Bartleby84]

if my Rev.B dual processor 2.0 is the base (i assume Rev.B) then that means that it is actually FASTER than the dualcore 2.0.???

The numbers I was quoting were just for the thread testing, and your systems should be comparable, if slightly faster. In the dual processor system, each processor is served by it's own independent bus, where as in the dual core system, both cores have to share a single bus. I'm not certain how much of an impact that would have on threads, but it's possible. Overall, however, the dual core systems are marginally faster than the dual processor systems at the same clock rate, primarily due to faster RAM and graphics cards, I think.

 

[MacEyeDoc]

Pie Is it Noisy?

All you lucky ducks with the new Quads:

How is the noise level compared to what you had before? I assume it is liquid cooled. There was a comment made somewhere that because Apple wasn't describing them as "whisper quiet" on their web site they were louder. Hope we don't have to go back into the "wind tunnel" to experience that 900% speed improvement. (Somebody tell me my math is off - could I really be able to purchase a 900% increment in speed?)

 

[EricNau]

Using Photoshop CS or CS2 (or PS 7)
==================================================
1.) Download the test image from http://www.quicklance.com/test.jpg
2.) Save it to the computer and then open it up in Photoshop
3.) From there please apply a ‘radial blur’ with the settings at:
Amount = 100
Blur Method = Spin
Quality = Best
Using a stop watch / ps timer see how long it takes to apply this filter
I just want to see what these new cpu’s can really do.

Results:
iMac G5 1.8GHz, 1GB - 2:00
Athlon XP3200+, 1GB - 2:15
Athlon64 4000+, 1GB - 1:25
Dual 2.5 Running 10.4.2 with 2.5 GB RAM 40 seconds
PowerMac Dual 2.7 Dell 2405 FPW, 2.5 gigs of ram, Radeon 9650 42 seconds
Dual Core 2.0 GHz G5 with 2.5GB ram Photoshop CS2 47.4 seconds

Quad 2.5GHz G5 2.5GB RAM 10.4.3 22 seconds !!!!! It's an amazing huh??!!!

Has anybody on a PC tried this? If so, I am very curious to hear the results. I would love to compare my iMac with my friends PC's, but I don't want to risk them beating me. What are your thoughts? Should I risk it?
My iMac (see sig) scored 1:50 using this photoshop test

 

[doodguy]

CINEBENCH 2003 v1
****************************************************

Tester : doodguy

Processor : Power Mac G5
MHz : 2.7 GHz
Number of CPUs : 2
Operating System : 10.4.3

Graphics Card : ATI 9650
Resolution : 1920 x 1200
Color Depth : 32-bit Color

****************************************************

Rendering (Single CPU): 384 CB-CPU
Rendering (Multiple CPU): 645 CB-CPU

Multiprocessor Speedup: 1.68

Shading (CINEMA 4D) : 362 CB-GFX
Shading (OpenGL Software Lighting) : 970 CB-GFX
Shading (OpenGL Hardware Lighting) : 1673 CB-GFX

OpenGL Speedup: 4.62

****************************************************

 

[aussie_geek]

Look at that memory fill rate - over 8GB / sec!!!

 

[AidenShaw]

don't worry about a nearly meaningless test

Has anybody on a PC tried this?

I would love to compare my iMac with my friends PC's, but I don't want to risk them beating me. What are your thoughts? Should I risk it?

Why should you care about losing a test of one Photoshop filter on one single image?

It doesn't really mean anything other than how fast you can radial blur the horse.

 

[nospleen]

My new imac got a 72.8. (i still have stock ram, my 1GB stick has not arrived.)

 

[Nikore]

I ran all three test on the desktop in my sig.

Xbench:

Results 117.66
System Info
Xbench Version 1.2
System Version 10.4.3 (8F46)
Physical RAM 4096 MB
Model PowerMac7,3
Processor PowerPC G5x2 @ 2.70 GHz
L1 Cache 64K (instruction), 32K (data)
L2 Cache 512K @ 2.70 GHz
Bus Frequency 1 GHz
Video Card GeForce 6800 Ultra
Drive Type HD
CPU Test 133.91
GCD Loop 135.73 7.15 Mops/sec
Floating Point Basic 136.23 3.24 Gflop/sec
vecLib FFT 136.13 4.49 Gflop/sec
Floating Point Library 127.92 22.28 Mops/sec
Thread Test 137.80
Computation 139.79 2.83 Mops/sec, 4 threads
Lock Contention 135.87 5.85 Mlocks/sec, 4 threads
Memory Test 106.18
System 103.85
Allocate 139.65 512.85 Kalloc/sec
Fill 133.13 6472.91 MB/sec
Copy 70.34 1452.86 MB/sec
Stream 108.62
Copy 107.24 2215.00 MB/sec [G5]
Scale 107.64 2223.77 MB/sec [G5]
Add 110.16 2346.74 MB/sec [G5]
Triad 109.51 2342.61 MB/sec [G5]
Quartz Graphics Test 133.97
Line 133.90 8.91 Klines/sec [50% alpha]
Rectangle 132.20 39.47 Krects/sec [50% alpha]
Circle 134.45 10.96 Kcircles/sec [50% alpha]
Bezier 129.42 3.26 Kbeziers/sec [50% alpha]
Text 140.38 8.78 Kchars/sec
OpenGL Graphics Test 149.17
Spinning Squares 149.17 189.23 frames/sec
User Interface Test 125.97
Elements 125.97 578.14 refresh/sec
Disk Test 75.50
Sequential 147.06
Uncached Write 199.56 122.53 MB/sec [4K blocks]
Uncached Write 182.43 103.22 MB/sec [256K blocks]
Uncached Read 85.46 25.01 MB/sec [4K blocks]
Uncached Read 199.73 100.38 MB/sec [256K blocks]
Random 50.79
Uncached Write 18.70 1.98 MB/sec [4K blocks]
Uncached Write 117.38 37.58 MB/sec [256K blocks]
Uncached Read 100.82 0.71 MB/sec [4K blocks]
Uncached Read 146.30 27.15 MB/sec [256K blocks]


CINEBENCH 2003 v1
****************************************************

Tester : Nikore

Processor : PowerMac G5
MHz : 2.7 GHz
Number of CPUs : 2
Operating System : 10.4.3

Graphics Card : 6800 Ultra
Resolution : 1600x1200
Color Depth : Milions

****************************************************

Rendering (Single CPU): 383 CB-CPU
Rendering (Multiple CPU): 666 CB-CPU

Multiprocessor Speedup: 1.74

Shading (CINEMA 4D) : 364 CB-GFX
Shading (OpenGL Software Lighting) : 1059 CB-GFX
Shading (OpenGL Hardware Lighting) : 1631 CB-GFX

OpenGL Speedup: 4.48

****************************************************


And lastly photoshop test (my stopwatch/photoshop's timer)

42.6/49.2


Conclusion: Quad > *

 

[gmnygts]

****************************************************
using Pie's numbers to calculate times(x) faster, using my machine as baseline;
Rendering (Single CPU): 1.255 faster
Rendering (Multiple CPU): 1.976 faster

Shading (CINEMA 4D) : 1.278 faster
Shading (OpenGL Software Lighting) : 1.281 faster
Shading (OpenGL Hardware Lighting) : 1.302 faster

glen

And the Quad G5 is 8.99 times (you could call it 900% faster) than my PowerBook. Wow! Gotta get me one of those![/QUOTE]

i believe that something is wrong with the CINEBENCH test that Pie ran

there should NOT be a 1.976 faster on the Rendering (Multiple CPU)
-- my approximation would be around 3.9 to 3.95 faster

glen

 

[EricNau]

Why should you care about losing a test of one Photoshop filter on one single image?

It doesn't really mean anything other than how fast you can radial blur the horse.

Quite Simply: If I were to lose to them, I would never hear the end of it, but I need a way to prove to them that my computer is faster. I just switched 6 months ago and all of them are convinced that Apple's are slow ("Only 2.0 GHx?").

 

[nospleen]

iGary, have you tried the photoshop test? I cannot believe the Quad did it in 19.2. That is just nuts!!!!