I've just bought some new Macs and I've found a few oddities with speed that I'd like to share... This is my old hardware:
- PowerBook 1.25Ghz, 1GB RAM
- PowerMac dual 2.5Ghz, 4GB RAM
And this is my new hardware:
- MacBook Pro Intel Core 2 Duo, 2.33Ghz, 3GB RAM
- Mac Pro, Quad Xeon 3Ghz, 4GB RAM
This is, as you can see, the very fastest desktop and laptop from Apple. I received the MacBook Pro first and decided I would pitch it against the PowerMac in a small test, since I felt that the MacBook Pro was really fast. So I downloaded the "mid=27:Scroll" Photoshop action from www.panosfx.com, which is a free download. It's a script that creates, from scratch, a parchment-looking image. After running this test, these were my scores:
| PowerMac | 46 seconds |
| MacBook Pro | 18 seconds |
Ok, so the MacBook, while executing non-native code is still over twice as fast as the desktop. That's pretty amazing! Rosetta rocks!. So I got my Mac Pro earlier this week. I thought I'd run the same test. So, the fastest dekstop against the fastest laptop from Apple. The results:
| Mac Pro | 53 seconds |
What? Are you seriously telling me that the laptop is almost three times as fast at executing this non-native test? Ok ok, relax. Maybe it's just that the Intel Core 2 Duo is better at running code in Rosetta, what do I know? Let's for the sake of sanity assume that it is.
Ah, but then Photoshop CS3 Beta was released. Well well well, now the tables are turning. Right? I mean, with a native Photoshop, it will totally scream. Well, sure it does:
| Mac Pro | 20 seconds |
Wow, now that's a huge improvement! Phew... Oh wait... Did you say 20 seconds?? 20? So, let me get this straight... My Macbook Pro runs the same action FASTER on a NON-native version of Photoshop than it runs on a NATIVE version on a supposedly faster machine?? What's up with THAT?? What will the benchmark be if I run CS3 on the Macbook Pro then?? Well:
| MacBook Pro | 9 seconds |
WTF? I mean, seriously. What is up with this? What is wrong with the Mac Pro? How could it possibly do this slower, with four cores on two 3Ghz Xeon Woodcrest chips?? I mean, am I in Twilight Zone all of a sudden? How how how is this possible? All the review of the Mac Pro has been about how damn fast it is. I'm seriously starting to wonder if this laptop isn't actually faster. Can someone else run the action on their Macbook Pros and their Mac Pros and report your speeds?
Xbench
If I pitch my two systems together with Xbench, I get this result, which shows the Mac Pro to be faster, but not by a lot...
If I pitch my two systems together with Xbench, I get this result, which shows the Mac Pro to be faster, but not by a lot...