Am I doing something wrong?
This is a new benchmark for me, and my times are terrible:
http://i650.photobucket.com/albums/uu225/AceNZ/pi.jpg
I ran the test with affinity set to one CPU and in realtime mode. My OS is Windows Vista Ultimate 32-bit.
Could the poor result be happening due to the 4MB of cache per CPU, vs. 6MB for some of the better results?
I do have a bunch of services running, but it's hard to imagine that being more than a 20% effect at the absolute maximum.
I know my memory timings are a little off. I've had trouble getting the numbers where they should be (the chips are spec'd at 5-5-5-15). My motherboard is a Gigabyte GA-P35-DS3P. Again though, this doesn't seem like a big enough difference to account for my results.
I have speedstep enabled, but it kicks up to 2400 MHz after only about 3 seconds. If I disable speedstep, the overall result only improves by about 1 sec.