No need to counter with stats, you partially debunked yourself! A linear copy test has no bearing on fragmentation's effects on real world random access. It wasn't 6 or 7 years ago, it was less than 4 this got hashed about. There's more "database" like activity than you think on the average PC and not all systems choose to use high RPM drives either.
Yes, if you use MyDefrag (which only guides the MS defrag engine) the results are much different. Good defrag has big impact on a machines hosting multi-GB games you WILL see a performance difference! Why? Because most all games use large archives of resources rather than individual files. Same is true of swap & hibernation files, especially on low RPM drives used in netbooks, because of the multi-GB nature of the file. I see the same with the 100GB+ of usenet headers the Newsbin Pro uses. The tests? I have 100's of GB in my game folder which Steam loves to update & patch which eventualy leads to persistent major stutter, defrag eliminates this. Same of excessive swapping or restoring from power save on my netbooks. On Wed, Aug 7, 2013 at 7:34 AM, Thane Sherrington <[email protected]> wrote: > At 09:16 PM 03/08/2013, Joshua MacCraw wrote: >> >> > >> > Because it's a waste of time and power? :) >> >> WRONG >> >> > >> > Have you ever seen a performance bump after you defragged? >> >> YES >> >> Been down this argument path before Thane... > > > Heh heh. Unfortunately, using the standard defragger in Windows, you are > incorrect. After a year of solid testing on literally hundreds of > computers, the different in normal file access before and after defragging > was less than 5%, an never went about 7%. At worst, that's statistical > noise, and at best is such a small change that it wouldn't be noticeable by > a human during normal use. > > The I did was test was done as follows: I ran a copy of all files on the > hard drive to nul to test file and timed it to get the read time . Then I > defragged and then re-ran the original test. > > My caveat: Using a good defragger (such as MyDefrag) which also optimizes > the drive properly can make a difference, and there are some applications > (such as databases) which will show a significant improvement by defragging. > > But for the vast majority of people, the vast majority of the time, the > speed up using the MS defragging program will never give a speed bump of any > noticable amount. > > Don't agree? Then give me a benchmark that proves your point. When this > was discussed (probably six or seven years ago), the above test I mentioned > was explained and no one came up with a better one. > > T >
