I have been setting up the Vista version of FRAGG and doing benchmarks, and I noticed that the benchmark results were fluctuating by up to 36%, whereas the WinXP results for the same test were all within a 16% range.
It took a while for me to realise that the delay of 200 seconds from startup to doing the test just wasn't long enough for Vista. It can take up to 5 minutes for Vista to settle down after a reboot. So I will have to re-run the tests with a boot delay of at least 300 seconds. What a pain, not to mention a delay.
The screen shot above shows how the Page File usage changes (thick blue line) as Vista gets going. The screen shot was taken at the point that I was loading the tests, and Vista is clearly still busy. Oops.
The delay in completing the results has been beneficial from another aspect: Tom West has been helping me tweak hIOmon to capture more meaningful file performance information. I haven't published the results of this data yet, because I think the meaningful info will come when we start comparing various products. hIOmon is an awesome product, and extremely powerful, but I'm getting old and RTFM isn't as easy as it used to be. Tom has been extremely generous in helping me get to grips with the product.
No comments:
Post a Comment