ALERT: According to Ladd at Xbench: 10.4 attempts to limit screen updating to 60 frames per second, and will throttle applications that exceed this limit. Normally, this is the correct thing to do, but for a benchmarking application it doesn’t work out so well. In the meanwhile, you can turn off beam sync in Quartz Debug.
They’re working on an update to resolve the issue. I’ll be posting updated benchmarks soon…
ALERT: According to Ladd at Xbench: 10.4 attempts to limit screen updating to 60 frames per second, and will throttle applications that exceed this limit. Normally, this is the correct thing to do, but for a benchmarking application it doesn’t work out so well. In the meanwhile, you can turn off beam sync in Quartz Debug.
They’re working on an update to resolve the issue. I’ll be posting updated benchmarks soon.
Following are some initial benchmarks comparing the PowerBook G4 1.5GHz (2GB RAM, 120GB HDD) and the MacBook Pro 2.0GHz (1GB RAM, 100GB HDD) from MacBench 1.2.
PBG4 1.5GHz | MBP 2.0GHz | % of | |
Results | 29.98 | 53.03 | 177 |
System Version | 10.4.5 (8H14) | 10.4.5 (8G1453) | |
Physical RAM | 2048 MB | 1024 MB | |
Model | PowerBook5,4 | MacBookPro1,1 | |
Drive Type | ST9120821A | ST9100824AS | |
CPU Test | 34.54 | 67.28 | 195 |
GCD Loop | 87.7 | 240.31 | 274 |
Floating Point Basic | 19.15 | 71.12 | 371 |
vecLib FFT | 85.16 | 38.17 | 45 |
Floating Point Library | 24.71 | 66.55 | 269 |
Memory Test | 19.4 | 104.12 | 537 |
System | 16.44 | 113.57 | 691 |
Allocate | 69.33 | 128 | 185 |
Fill | 15.14 | 109.51 | 723 |
Copy | 9.81 | 105.57 | 1076 |
Stream | 23.65 | 96.12 | 406 |
Copy | 21.15 | 81.4 | 385 |
Scale | 24.66 | 80.12 | 325 |
Add | 24.85 | 119.82 | 482 |
Triad | 24.36 | 117.64 | 483 |
Quartz Graphics Test | 45.77 | 45.19 | 99 |
Line | 38.73 | 46.65 | 120 |
Rectangle | 44.22 | 40.44 | 91 |
Circle | 47.31 | 41.4 | 88 |
Bezier | 55.78 | 74.07 | 133 |
Text | 45.99 | 37.27 | 81 |
OpenGL Graphics Test | 65.71 | 129.91 | 198 |
Spinning Squares | 65.71 | 129.91 | 198 |
User Interface Test | 19.6 | 23.97 | 122 |
Elements | 19.6 | 23.97 | 122 |
Disk Test | 21.08 | 32.45 | 154 |
Sequential | 38.04 | 58.02 | 153 |
Uncached Write | 33.79 | 71.99 | 213 |
Uncached Write | 33.82 | 51.13 | 151 |
Uncached Read | 49.35 | 47.79 | 97 |
Uncached Read | 38.9 | 68.65 | 176 |
Random | 14.58 | 22.52 | 154 |
Uncached Write | 4.71 | 7.29 | 155 |
Uncached Write | 39.89 | 66.61 | 167 |
Uncached Read | 49.45 | 69.24 | 140 |
Uncached Read | 59.63 | 90.79 | 152 |
Note:
PowerPage reader dombi notes that there’s a bug in Xbench when running on Intel. You need to Turn off the Beam sync option within the Quart Debug application found on the developer tools). Better MacBook Pro benchmarks can be found here.
4 replies on “MacBook Pro Benchmarks (Update2)”
To the photoshop guy – O’grady has a basic benchmark of Photoshop at https://www.powerpage.org/archives/2006/02/the_apple_core_macbook_pro_photoshop_benchmarks.html
Basically the performance of Photoshop on his 2 tests falls roughly half way between a TiBook 1Ghz and a PowerBook 1.5Ghz… closer to the PB than TiB.
He links to the ZDNet article he published – I’d give you that but this whole “no free link advertising” got me worried (and the “Terms” don’t come up when I click ‘terms’!)
Hope that helps!
There is a bug in Xbench when running on Intel. You need to Turn off the Beam sync option within the Quart Debug application found on the developer tools).
Here are the correct MacBook Pro results:
http://db.xbench.com/merge.xhtml?doc1=157247&doc2=1&setCookie=true
My CPU Test yielded 69.15 overall (139.69 GCD, 46.48 FP Basic, 98.27 FFT and 52.65 FP Lib) – 1.67Ghz Powerbook.
Compared to above as 100%, the MBP has the following:
Overall: 97%
GCD: 172%
FP Basic: 153%
FFT: 39%
FPLib: 126%
I’m confused Erik, how come your PB G4 1.25 significantly outperforms a PB G4 1.5? Is it the same test?
Looks like Quartz is NOT optimized for Intel yet. From what I’ve read other places, I guess I should be surprised it even runs. The lines that surprise me most are “Copy 9.81 105.57” and “vecLib FFT 85.16 38.17”. Both of them seem very weirdly skewed…in opposite directions. But I don’t even know what they mean.
Other than that, it looks safe to say that the machine is generally about twice as fast, which should result in a noticeable improvement in speed during use. Is this the case?