The benchmark application only uses one processor, but OSX using both. This means that while one processor was doing the benchmark, the other was playing an iTunes song. One good way to see what your processor is doing is to open up the CPU monitor utility. This just shows total CPU usage. The Process Viewer utility shows in more detail what the processor is being used for. If you change the sampling rate from the default 20 secs to 5, you can get a better idea of what is going on. Run the benchmark while you are watching the Process Viewer, and you will see that every once in a while, the Login process appears. Also you see that at certain times, the Window Manager (Aqua?) jumps up to about 10-15% and then back down to normal 2-3%. I think both of these will be changed in the new version 10.1, and we will be able to see a slight increase in the times for this benchmark.
In case anyone is interested I ran a few benchmarks on my not so new dual 533 ( digital audio, no fancy codenames for me =( )running OS X 10.0.4, 256 Megs. of RAM, 7200(?) RPM stock 20 gig. ATA, Adaptec Ultra160 controller attached to a 10k RPM Quantum 9 gig. and a nVidia GeForce2 MX.
ok, here goes...
Quickbench Time: 1.38333332 (min.)
- i beleive it uses both CPUs indirectly cause i had other programs open at the time but none using the CPU, and while it was running CPU monitor showed both CPUs jumping up and down.
Altivec Fractal: 3351.4 MegaFLOPs (4.0 sec)
On the lighter side everyone seems to complain about OS X being slow. In X my DP is only a little bit slower, but greatly makes up for it with the multithreaded system. Ill be happy to take the speed hit considering it all even out when you add all the hangs you can run into in 9 while awaiting something to render/get connection/etc....