Quote Originally Posted by cbjaust View Post
do you find the Sandra memory bechmark to be consistent? I have never found it to be consistent at all; running consecutive runs of that benchmark I would see low numbers initially that then climb to a mamimum but with so much variation it's hard for me to trust the numbers. then again I always run it unbuffered and with the optimisations off.
Cheers
You are right that benchmark has a huge variation between runs. What do you mean by running it unbuffered? I use fusion to disable unneccessary background tasks that helps abit.