If two disks have the same capacity, but different organization (say 4 sides and 614 cylinders versus 8 sides with 307 cylinders), is there any useful difference between them?
-
What exactly is your question?– fretjeDec 7, 2009 at 20:41
-
Here's a paraphrasing: "If I had two (say) 500GB drives and one has more physical platters than the other, could the drives perform measurably different?"– Chris W. ReaDec 7, 2009 at 21:02
-
4@fretje: "Is there any useful difference between two disks with same capacity but different geometry?"– quack quixoteDec 7, 2009 at 21:02
-
@BasicallyMoney.com: that's a fairly specific interpretation of "useful difference"; the OP may be thinking of another interpretation. (it is rather vague, isn't it...)– quack quixoteDec 7, 2009 at 21:04
-
I am thinking of the performance , when doing a seek from one cylinder to another But I am not quite sure any help will be appreciated.– Smart GirlDec 7, 2009 at 21:26
2 Answers
Most (all?) modern hdd's geometries are 1) fictional and/or 2) irrelevant these days. There was a time when the OS did make useful decisions based on the hdd geometry, but these days, the fiction of c/h/s is managed by a controller internal to the hdd, and most (all?) OS these days issue SCSI/ATA commands that treat the device as a flat array of blocks (LBA).
This means that the OS is not likely to be able to optimize performance based on these characteristics, but you can still probably measure performance differences with different geometries, if you setup the right kinds of tests.
These days, I mostly only pay attention to 1) spec (e.g., 3GBps SATA [a.k.a. SATA Rev 2.5]), 2) rotational velocity, and 3) on-disk cache size
I would think more platters would decrease performance, considering several reasons listed below in the links, but then it could be the opposite if the HDDs were from a different manufacturer each.
Check out these two links, if anything an interesting read: