Comparing Apples and Lemons*

So how do you compare an Apple running OS X with a Windows based PC? Many a flame-war has started over this matter. Just a few examples:
Win vs OSX

  • Apple geeks will tell you any PowerPC 0wnz Intel. But it seems Intel 0wnz Apple soon.
  • Microsoft alledgedly has copied the Mac interface, in fact it was a Xerox first. (On a sidenote: How come I can spend a day just clicking around on Win, while having to resort to keyboard on OSX? Just stick with “the Aqua eye candy is no match for the Luna Theme“.)
  • Apple hardware is more expensive than your generic Dell. Well, that is a fact, but the jury is still out on on the verdict of lemons.

In short: The MS camp stands squarely at the Apple camp. And vice versa.

Back to the question: How do you compare an Apple running OS X with a Windows based PC?

Simple answer: Benchmark.
Advanced answer: Take two or more pieces of hardware. Run several, industry standard tests. And run them several times, under the same conditions. Conditions like room temperature, no interference of background processes, etc. The tests have to be repeatable by outsiders. Take the test measurements and compare them using acknowledged statistical methods. The results will tell which one is the better (overall).

Dell C840 vs iBook G4

Enough with the intro already. I currently use a Dell Latitude C840 1.8 Ghz (Work) and an iBook G4 1.33 GHz (Leisure). All I want to know. How do they compare? Sure, I bought the iBook G4 12″ just for browsing, and it’s nice and small. It seems a capable machine altogether. Just how capable? Could it beat my ‘trusty’ Dell workhorse? (If you want the exact specs, just ask)

Enter the benchmarks. What kind of ‘industry’ standard tests could I run on Win AND Mac? This got me thinking. What DIY benchmarks are available on Win AND Mac? Here’s my current list of cross-Windows/Mac OS X applications:

  • Photoshop: This probably will have to do some ‘hard’ filter stuff on the same image, timing it. A nice indicator of FPU/Altivec/SSE performance.
  • Games: Quake III & UT2004 are available on the Mac as well. Just need to find out how to benchmarks with them. (If any game developer wants to supply a Win/Mac test suite, I’d be happy to test.) Timing FPS is a good indication of CPU/GPU performance.
  • Dreamweaver: Anyone’s favorite WYSIWYG HTML editor. Maybe could do some Office documents cleaning, as this is notoriously slow. (Thanks to MS’s proprietry markup.)
  • OpenGL: This could be a test of the GPU. The Dell has a nVidia GForce Go 440 and the iBook an ATI Radeon 9550. The latter one probably wins because of the FPS-race. Fact, there is a standard test-suite for OpenGL, somewhere.
  • Webserver: Apache & PHP is available for OS X Tiger natively. On Windows there are packages like XAMPP. Probably need to do some hard PHP work like looping a concat(?). MySQL is officialy slow on Mac OS X, so no need testing that. (Dutch).

So, I plan to run these tests on my machine and see how they compare. Keep an eye on this site. 🙂

Things not to test: Office, how fast can you type and apply a style? Excel maybe, if it’s a 100.000+ pivot-table excercise. The key is: ‘different hardware, same software, but pushing it to the limit’.

*Actually it is ‘comparing Apples and Oranges‘. Pun intended.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.