AquaMark3 has been a well-known Gaming benchmark for several years. And the fact that it even runs in this environment impresses me. That being said, if you want to run PC Games in Parallels, even with Direct X 8.1 (and not 9 or 10,) the performance needs to be improved. AquaMark3 can be downloaded here: AquaMark3 I ran it in Windows XP Home (Boot Camp) in Parallels (Full Screen w/64MB VRAM) in 1024x768 mode on my MacBook Pro 2.16GHz ATI X1600 machine. It got an average of 4.2 FPS and a score of around 4,200 in Parallels. I also noticed that Parallels seems to be restricted to only 103% CPU (with Duo Core the upper limit is 200%.) If Parallels allowed more CPU usage, this number could be greater. There were also a couple of Graphical Artifacts (errors) not seen when running Windows Natively. They should be pretty obvious to anyone repeating this procedure. (Negative images/stripes.) In order to play a 3D rendering game, this really needs to be around 30,000. That's approx. 30 FPS. When in Boot Camp, my MacBook Pro gets an AquaMark3 score of about 32,709, or an average of 32.7 FPS. I realize that that is an absolute max. that Parallels can only run a certain percentage of, but with optimization of your technology, I would think you could get to > 20 FPS. That would make gameplay on Direct X 8.1 games acceptable (IMHO.) If Parallels could perform this benchmark well, MANY PC games that use Direct X 8.1 would also run well.
The reason that the CPU is capped is because Parallels uses a single core of your MBP versus both, that way you have adequate OS X preformance. I'll check back with some benchmarks on my laptop, I thought I got more than 4.2 FPS when I ran them before. Finally, if Parallels did well on this benchmark, they don't neccesarily run the games better, right? From the classic syllogism: All men are mortal. Socrates is a man. Therefore Socrates is a mortal. All men are mortal != All mortal things are men. Not to be a pain, but yeah. Just a pet peeve of mine.
I pretty much got the same results you did... I'll go grab another benchmarking utility and see how that does.
Part of the reason I used AquaMark3 is that I HAVE tested it on many platforms, and those that did well with this benchmark, also did well with game graphics. That's why this benchmark was one of the big three, used by Tom's Hardware, back when DirectX 8.1 was released. It also does a good job of evaluating system performance for games that use DirectX 9. I used it to evaluate several systems for Star Wars Galaxies, and each system that had a score over 30,000 got adequate performance from the game.
