The Math Behind Short Stroking HDDs

M

Mr. K6

Supreme [H]ardness
Joined
Mar 23, 2005
Messages
5,077
I ordered two new 1TB WD Caviar Black's to replace my current two-year-old 320GB Seagates. I'm excited for the performance increase but more so want to make sure I get the most performance out of these new harddrives. I've been short-stroking my hard drives for awhile, but was wondering if anyone had ever mathematically figured out the optimum configuration/percentages/set-ups for short-stroking. Googling has turned up nothing.

If not, I imagine it would be relatively simple to do some quick calculus to the figure out what would be optimal. Basically, one would figure out the rate of speed penalty as one moves from the outer edge of the platter to the innermost edge. Next, one would figure out how much space one gains moving from the outer edge of the platter into the innermost edge. Analyze the curves and you have your answer of how large of a partition to create. However, it might not be so simple, and I don't have the background in HDD technology to verify my hypothesis. First, is HDD space directly proportional to area across the platter? Or are there more dense zones and less dense zones? Also, is there anyway to make sure that you're writing to the outer edge of the disk with your first partition, or does Windows do that automatically, every time? Is there a program where you can actually "see" where your partitions physically are on an HDD?

If my hypothesis is correct, I want to write up some quick equations based on proportionality, and then one can easily multiply by platter density and number of platters to figure out the final "optimal" partition size for maximum speed (or where and how much give and take you can have).

Thanks :cool:
 
The optimum is just using it all. Hell, just make different partitions if you're concerned about it. Put the files you don't need fast transfer speeds on the inside of the disk (which would mean the last blocks, since hard drives work from the outside in)
 
I would use it all, buying the HDD and then only using the short-stoked partition would be a waste. However, I would, say, install the OS to the short-stoke partition and leave media files on the rest of the HDD, i.e. use the speed where you need it. And how would one "put files" on the inside of the disk? Is there a specific program that allows you to do this? Also, wouldn't creating partitions like I originally planned be better since it would prevent the spread of the data away form their optimal positions due to fragmentation?
 
I like where this thread is going.

The disk sectors are numbered sequentially and start on the outside of the disk, so if you make a small partition at the beginning of the drive, it will be on the outside.

Data density is pretty much constant across the whole disk, and you can easily see the performance curve with apps like HDTune.

So, if you did an HD tune test on a drive, then took those results and plotted them reasonably into excel you could then come up with another curve for data density and setup a nifty little graph that would show you the optimal setup.

Although, you would need to know some physical stuff about the drive, like the number of tracks, and the amount of data on the innermost track and the amount of data on the outermost track, and that information I doubt is public. Perhaps if you just went with physical dimensions (outer diameter, inner diameter, and then did some math to divide the total amount of space by the number of platters) you could make a reasonable approximation....

Sounds kind of fun actually...
 
Reasonable approximation with HDTune is actually exactly what I have been doing :D. For instance: http://www.gamingshogun.com/Article/3623/Western_Digital_Caviar_Black_1TB_Hard_Drive_Review.html

Looking at that, I estimate no to very little performance loss up to about ~10%. Now is that in relation to the whole drive or a single platter? I mean, HDD's write to each platter at the same time as they move in, not like, they fill up part or most of the first platter before going to the second, right? (I would hope so, man that would be inefficient if they didn't) So, that said, up to ~10% means I could create a 100GB partition on the 1TB WD Caviar Black that would be extremely quick (not 33GB, right?)

The other thing I'm throwing around is whether or not it would be worth it to do the performance curves. My reason here is that the calculations and formulas would only be as accurate as the data. If the data from say HDTune or physical measurements has a large enough percent error, then it might not be worth it to do more than estimate/extrapolate from HDTune tests.
 
I'd imagine that the ideal settings would vary with the specifics of each drive model. I think the best way is just to run HDTune and see where the performance level drops off to what you consider unacceptable, and make a partition matching that size.

The geek in me would be interested in seeing the math, just because I like to see how things work. However, it seems like a bit of going backward to go forward - you're taking the performance numbers to figure out the specs to figure out the ideal performance numbers. Seems like it'd be just as easy to just use the first performance numbers.
 
First off I want to point out that density is consistent across the platter, because if it were not that could cause some serious issues.

Secondly SS is only useful if you want to physically separate OS and DATA.
Also SS will only help so much, and it really only shows up in synthetic benches.

Yea over time it will keep your OS on this one part only but the differences will be minuscule and not noticeable by an end user.
 
You must log in or register to reply here.
Back
Top