1.67ghz G4 vs. 2ghz Pentium M: Its official

MacTruck

macrumors 65816
Original poster
Jan 27, 2005
1,242
0
One Endless Loop
Ok, so I have the following computers:

1.67ghz Powerbook G4 128mb Vram 5400 100gb HD

IBM Thinkpad 2ghz Dothan 2mb cache Pentium M 60gb 7200rpm HD 128mb VRAM Fire GL video

I downloaded the 1080HD BBC video from apples website and ran it on both systems. Both systems play the video choppy, infact the G4 plays it a little less choppy. The Pentium M ran at 100% cpu and choked the whole way. Now compare these 2 systems. The G4 has a 167mhz bus and the IBM has a 400mhz bus. Other than that the systems use the same memory, have the same video card basically and have the same amount of ram that runs at the same speed pc2700. The IBM even has a faster Hard Drive. Now keep in mind that Dothan in the IBM has a 2mb cache and the G4 has 512 on chip. You would think the Dothan would come out ahead and play that video no problem but it won't.

So there you go folks, the Pentium M is no faster than the G4. Intel better have something alot faster for us in 2 yrs.
 

mad jew

Moderator emeritus
Apr 3, 2004
32,194
6
Adelaide, Australia
You used Quicktime to play the videos, right?

I've found my 800MHz iBook is faster with iTunes than my 3.0GHz PC but that's simply a matter of optimisation. Could this explain some of the difference?
 

MacTruck

macrumors 65816
Original poster
Jan 27, 2005
1,242
0
One Endless Loop
mad jew said:
You used Quicktime to play the videos, right?

Yes, both were using the latest Quicktime 7 software. Not 7.1.

Another note. When playing I tried to resize the window in both computers. The mac allowed resizing no problem while the Dothan CPU choked when I tried to resize the video window. The G4 is just faster.

The only flaw in this test is if apple purposely made the windows version of quicktime slower somehow or if XP plays a factor in the speed. I doubt it though.
 

WinterMute

Moderator emeritus
Jan 19, 2003
4,715
4
London, England
MacTruck said:
Yes, both were using the latest Quicktime 7 software. Not 7.1.

Another note. When playing I tried to resize the window in both computers. The mac allowed resizing no problem while the Dothan CPU choked when I tried to resize the video window. The G4 is just faster.

The only flaw in this test is if apple purposely made the windows version of quicktime slower somehow or if XP plays a factor in the speed. I doubt it though.
Both systems use software to decode the video, there's your problem, XP isn't particularly good with video anyway, plus QT is a core element of OSX and an application level system in XP, they handle the data differently.

I don't see how this can be a fair test, as the operating systems are different, it will be clearer when you can run the test with a Mac/intel machine running the same OS.

I'm pretty sure the M version that Apple will get won't be your standard Pentium M anyway.

I'm always very sceptical of cross platform testing that doesn't involve calibrated benchmarks, way too many variables.

Still, it's good to know the current PowerBooks are competitive in the field.
 

jadam

macrumors 6502a
Jan 23, 2002
699
1
MacTruck said:
Its hardly useless. Got a better idea for a test? I will run it.
No, it's definetly useless.

Trying doing a mathematical test.
 

keysersoze

macrumors 68000
Jan 6, 2004
1,596
8
NH
edesignuk said:
This is a next to useless comparrison. Apple make QT, of course it runs better on OS X :rolleyes:
haha, I was about to write the same thing, but e beat me to it. I am sure some better tests are out there that would allow for more of an Apples-to-Apples comparison.

I would expect the pentium-m to kill the G4.
 

MacTruck

macrumors 65816
Original poster
Jan 27, 2005
1,242
0
One Endless Loop
keysersoze said:
haha, I was about to write the same thing, but e beat me to it. I am sure some better tests are out there that would allow for more of an Apples-to-Apples comparison.

I would expect the pentium-m to kill the G4.

You can't get more apples to apples than the same software on both systems with the same specs. I think you will be dissapointed in the Pentium M.

No matter what test is done someone will always say its flawed. I bet in 2 yrs when apple is running on intel you guys will say that OS X on intel is not optimized yet right. LOL.
 

mad jew

Moderator emeritus
Apr 3, 2004
32,194
6
Adelaide, Australia
MacTruck said:
You can't get more apples to apples than the same software on both systems with the same specs.

It was a good idea MacTruck but the problem is the software part. It's not the same software on both computers despite having the same name and coming from the same company.
 

drison

macrumors regular
Apr 30, 2005
168
0
Chicago
MacTruck said:
You can't get more apples to apples than the same software on both systems with the same specs. I think you will be dissapointed in the Pentium M.

No matter what test is done someone will always say its flawed. I bet in 2 yrs when apple is running on intel you guys will say that OS X on intel is not optimized yet right. LOL.
Completely different operating system, driver architecture, etc. It might be a fun comparison to make but hardly scientific.
-Dave :)
 

MacTruck

macrumors 65816
Original poster
Jan 27, 2005
1,242
0
One Endless Loop
Ok here we go.


Both systems running photoshop 7.0.1


- Create new image size 640x480 72dpi
- Fill with default black
- Filter Difference Clouds

Here is the timed test:

- Change image size from 72 dpi to 1600 dpi


Powerbook G4
- Finished in 11 seconds

IBM THinkpad
- Finished in 68 seconds


Now thats as even a test as you can get.

G4 way faster.
 

JFreak

macrumors 68040
Jul 11, 2003
3,145
0
Tampere, Finland
MacTruck said:
Got a better idea for a test?
the only reliable test will be one that is using exactly same software. that is, those universal binaries that share exactly everything except have different compilers targetting different cpu architechtures. this is something that has never before been possible, and apple has opened a possibility of direct and fair comparison between PPC and x86.

imagine fastest intel-inside macintosh ever (the developer one apple rents) and fastest-ppc-mac-money-can-buy running same natively compiled apps. whatever benchmark programs will be compiled will tell us the truth, given that both compilers offer somewhat similar level of optimization.

you could guess that such direct comparison would drive r&d up on each competing architechture, and i believe that is one of apple's intentions with this transition. now we will really see if the x86 sucks or not.

so hold your horses for a while. you just cannot compare quicktime-on-windows to quicktime-on-osx to get fair results between x86 and ppc performance. that is absurd. if x86 loses it is surely because windows sucks so bad. you will have to wait until you are able to compare quicktime-on-x86-osx between quicktime-on-ppc-osx to get fair results, and there is no other way. only reliable comparison is the one where is exactly one variable.
 

mad jew

Moderator emeritus
Apr 3, 2004
32,194
6
Adelaide, Australia
MacTruck said:
Denial man. Plain denial.

Nah, you're comparing two different programs - too many variables.

If you really want to see which computer is faster; hold them vertical and drop them out of a two story building. The Apple is more aerodynamic so I'm betting it'll reach the ground first. :)
 

nrd

macrumors member
Mar 8, 2005
83
0
New Jersey
MacTruck said:
Ok, running a photoshop test on both systems. Both systems using photoshop 7.0.1. Results momentarily....
You're still not doing an even comparison. OS X and Windows still factor in.

An even comparison would be using a Ubuntu Live CD on both your Powerbook and Thinkpad and testing using applications bundled with the CD. There's an x86 and PPC version of the live CD. Using them will not affect what is on your hard drive. That's as close as a one-to-one test as you're likely going to get.

You can use the GIMP as a replacement for photoshop.

http://www.ubuntulinux.org/
 

nrd

macrumors member
Mar 8, 2005
83
0
New Jersey
MacTruck said:
Denial man. Plain denial.
Unless you perform a truly even comparison between the two, you're living in denial. Might as well proclaim "Mission Accomplished" and close the thread.

Seriously, run Ubuntu Live CDs on both machines and prove the naysayers wrong. What do you have to lose (besides the time performing the test) ?
 

MacTruck

macrumors 65816
Original poster
Jan 27, 2005
1,242
0
One Endless Loop
Never heard mac faithfulls talking this much bull about the photoshop benchmarks when apple did them. Now that mac is going to intel the tides have changed. Face it, intel is not faster than PPC and won't be for a long while.
 

jsw

Moderator emeritus
Mar 16, 2004
22,817
37
Andover, MA
There is absolutely no way to ever directly compare the PPC and x86 chips in any meaningful way (i.e., apps), because there's always some different hardware, there's always a different OS - even with the Developer Transition Kit Macs, the x86 version of OS X is, at the bottom, different than the PPC version.

I think the Photoshop comparison - if the hard drives and RAM are equivalent - is reasonable, as they are what a user would experience, which ultimately is all that matters. However, that still doesn't do anything as far as showing that a G4 is or is not better than a Pentium M.
 

decksnap

macrumors 68040
Apr 11, 2003
3,070
70
It IS indeed more of a test of X vs. XP than it is hardware vs. hardware. But it does tell us SOMETHING. Either XP is horribly poor, or we're in for a world of hurt. My guess is somewhere in the middle.
 

ewinemiller

macrumors 6502
Aug 29, 2001
445
0
west of Philly
MacTruck said:
Its hardly useless. Got a better idea for a test? I will run it.
How about cinebench, and make sure to post the OpenGL and render results seperately that way if one has a really good video card and one has a crappy one, we can look at render performance.

Also make sure that you've got your thinkpad plugged in and don't have some whacky power saving mode taking over. I'm really surprised by that photoshop test. I have a dell Pentium M and an ibook G4. I don't do the photoshop thing, more towards the 3D rendering and Pentium M is faster than the G4 clock for clock on that.