HDD Speed

[Sc4freak]

As I understand it, the outer tracks of a hard drive are the fastest. I recently bought a new Western Digital 500gb drive, and I copied ~200gb of data over to it. I noticed that the speed increased as the copy progressed, it started at about 20MB/s and increased to almost 40MB/s by the end of the copy.

I had always thought that hard drives wrote data from the outside in, but it appears not. If this is the case, doesn't that mean that when I install my OS, it'll be located on the slowest part of the drive (ie. the inner track)?

Is there any tool that I can use to confirm or debunk my findings?

 

[The_Wolf_Who_Cried_Boy]

Are you sure you weren't being limited by the source disk's read speed?

 

[Davros]

There is but i cant for the life of me remember the name

edit:
could be paragon hard disk manager
there is a free trial available
http://www.paragon-software.com/home/hdm-personal/download.html

 

[The_Wolf_Who_Cried_Boy]

Ultimate Defrag lets you put files where ever you want.

I've never used it but I've seen good things said about it.

 

[Tokelil]

You could run HD Tune to make sure that it reads from the outside in:
http://www.hdtune.com/

I have never seen a disk that didn't though. (As you said, it would throw of all the optimizations etc. the OS does regarding HDD speed.)

 

[mito]

Ultimate Defrag lets you put files where ever you want.

I've never used it but I've seen good things said about it.

Interesting. Anyone tried this one out?

 

[swaaye]

I wonder if that could screw up access times though. You would probably end up with files all over the drive and make the head have to go farther more often.

 

[Mintmaster]

I wonder if that could screw up access times though. You would probably end up with files all over the drive and make the head have to go farther more often.

Movement time for the read head is only part of the total access time, and access time is only part of HDD performance. 10K RPM drives like the Raptor aren't much faster than top 7200 RPM drives nowadays, despite a much lower access time and faster transfer rate.

http://www.storagereview.com/HDS721010KLA330.sr?page=0,1
http://www.storagereview.com/HDS721010KLA330.sr?page=0,2

 

[Frank]

AFAIK, except for the old Apple Superdrives and linear track media like CD and DVD, everything else has the same amount of sectors for each cylinder no matter if it's on the inside of outside of the platter. The bits on the inside are just much smaller.

 

[Tokelil]

AFAIK, except for the old Apple Superdrives and linear track media like CD and DVD, everything else has the same amount of sectors for each cylinder no matter if it's on the inside of outside of the platter. The bits on the inside are just much smaller.

How do you explain results like ie.:

 

[Frank]

Well, how would you explain that nice, gradual slope? Remember, most disk nowadays always rotate at the same speed.

 

[Tokelil]

Well, how would you explain that nice, gradual slope? Remember, most disk nowadays always rotate at the same speed.

Yes, so the slope is there because the transfer speed is faster at the outer tracks than the inner, because each bit is the same size and the outer track contains more bits than the inner track. (Time for each rotation is the same, but the velocity at a track depends on the distance to the center.)

If each track contained the same amount of sectors, then the transfer rate would be linear over the surface, which it clearly isn't.

 

[Dominik D]

Please note that defraging with 3rd party apps usually slows down machines using windows system recovery.

 

[Albuquerque]

AFAIK, except for the old Apple Superdrives and linear track media like CD and DVD, everything else has the same amount of sectors for each cylinder no matter if it's on the inside of outside of the platter. The bits on the inside are just much smaller.

Even CD's and DVD's read faster from the outer track, unless you're using a really old CLV drive. Basically every CD and DVD player in production today is a CAV device, which just like hard drives, means the outer edge will always read faster than the inner edge.

And why would they make "smaller bits" only to be used on the inside edge? That's nonsense at best... By the "smaller bits" logic, they'd STILL be stupid not to take those same smaller bits and put them all over the drive to get back another huge chunk of storage space on the same platter size.

There's nothing logical about putting "fat bits" on the outer edge and "thin bits" on the inside edge. The smaller you make ALL of them, the more data you can stuff onto that platter.

Note for anyone who is confused by the CAV/CLV nomenclature:

CAV = constant angular velocity, meaning RPM is constant (datarate changes based on position of read head in relation to platter size)

CLV = constant linear velocity, meaning datarate is constant (RPM changes based on position of read head in relation to platter size)