Sign in with
Sign up | Sign in
Your question

hard drive I/O question and comparison

Last response: in Storage
Share
August 16, 2008 6:29:25 PM

i plan on building a new PC and was doing performance comparisons on hard drives, i was looking over the benchmarks tom did here http://www.tomshardware.com/charts/hard-disks/average-read-transfer-performance,658.html and had a couple questions.

the Segate Barracuda 7200.11 placed almost as high or higher as the VelociRaptor in almost every benchmark except for the Database I/O Benchmark Pattern and the Fileserve I/O Benchmark Pattern benchmarks where it finished almost dead last for both. i was wondering how much of a performance cutback this is or what exactly this benchmark means.

also the Segate drive runs about 4 times cheaper than the WD drive so would a Raid 0 of 2 or more Segate drives out perform the VelociRaotor?

August 16, 2008 6:37:03 PM

It will depend on what you are doing. For the average user you would do far better putting the money towards something else in your system than a fast hard drive. In terms of perceived system performance RAID 0 will do little, although for sequential transfers of large files you'll get a big boost.

-mcg
a c 167 G Storage
August 16, 2008 7:14:00 PM

There is generally no real world(vs. synthetic transfer rate benchmarks) performance advantage to raid of any kind.
Go to www.storagereview.com at this link: http://faq.storagereview.com/tiki-index.php?page=Single...
There are some specific applications that will benefit, but
gaming is not one of them. Even if you have an application which reads one input file sequentially, and writes
it out, you will perform about as well by putting the input on one drive, and the output on the other.

Go to their benchmark database and look at the office drivemark tests for an indication on how different drives will fare in a realistic single user environment. http://www.storagereview.com/php/benchmark/bench_sort.p...
Related resources
August 16, 2008 7:35:40 PM

the main question im asking isint about will the raid make a big difference, but i really would like to know between these 2 drives what difference will the I/O benchmark have over regular computing. im already investing into othe good pieces, E8600 processor and 4870 x2 video card so basically just wondering about the hard drives.
a c 167 G Storage
August 16, 2008 9:42:57 PM

The synthetic benchmarks are indicators of performance, but they mean squat, in real applications.
One reason is that they are run on clean new drives, and only use the fastest part of the drive. In this case, larger drives may be good, assuming you are prepared to only use 10% of the capacity.
Next, the I/O patterns in benchmarks are unlikely to bear any resemblance to what YOU will be doing.
The server type benchmarks are meaningless, because a single user will rarely have a queue length over 1 making NCQ queueing useless.
Servers are characterized by many small random reads, while most users will have more sequential processing.

If you have the funds, I suggest a 300gb velociraptor. You can feel the difference, but it is not dramatic. Add a larger ,slower, storage device when you need more than 300GB.
!