Page 1 of 1
Hard Drive Information
Posted: Sat Jan 20, 2007 8:15 am
by dzjepp
Have a glance at this pdf (it comes with the ultimatedefrag app). Skip to 'hard drive performance theory'.
http://www.uploading.com/files/5YONX1JM ... e.pdf.html
How much of that is true? Do the tracks leading from the most outer to the most inner vary in speeds? (fastest to slowest). And in such when creating partitions on a drive, say 4 partitions, are they stacked in tracks from inner to outer? (Meaning c partition would be fastest, f partition the slowest).

O_o
Posted: Sun Jan 21, 2007 1:15 am
by PieceMaker
Hmmmmm.
Not sure on how much is true.
I have no help for you on this but I've be thinking of trying this myself.
*waves* lo dz!
Posted: Sun Jan 21, 2007 2:56 am
by ^misantropia^
Disclaimer: it's been a few years since I last attended hardware design classes and technology probably has changed a bit. With that in mind...
- Think of a wheel and how the rim spins way faster than the hub. Same with a hard drive. Doesn't really influence data transfer rates though because the nearer you get to the rim, the more data gets spread out over the platter (otherwise the head can't discern the individual bits).
Lots of R&D has and is being done on this subject, though, so perhaps these days the rim does sport the higher transfer rates. OTOH, there's also been a lot of research on squeezing more bits into each unit of physical drive space, which could mean the regions near the hub have the highest transfer rates since the platter moves slower there, giving the head more time to dissemble the information. YMMV.
- The middle of the drive is the best position to store often accessed data because, on average, the head will be closer to the center than to the edges, thus seek times will be shorter. Only a rule of thumb though: if the file you want resides in sector x, and the head is at (x + 1), the disk first has to make a whole spin. You'll understand what that does to performance.
Posted: Sun Jan 21, 2007 4:03 pm
by plained
i been doing this for years
outer rings for system or anything speed needy .
its what i doen sinse my amiga recording days, it was the only way i could record steady without the computer failing to provide the speed
i also did it for my maps they loaded real fast
almost always first there hehe
Posted: Sun Jan 21, 2007 11:51 pm
by dzjepp
Thanks yo guys.
Have you looked at the pdf btw? There's some interesting stuff in there.
Posted: Mon Jan 22, 2007 12:17 pm
by ^misantropia^
Nope. The site, apart from redirecting me to passion.com when I enter the code, doesn't do much for me.
Posted: Mon Jan 22, 2007 4:35 pm
by Underpants?
Disk performance seems to be a lot of voodoo and hearsay when applied to most modern disks.
Rather than physical location (ie, closer to the hub vs. farther away), using diskpar (along with diskpart for 2003) to align tracks relative to zone bit recording is where true performance (in windows) lies (up to 20% performance increase in some applications).