|
Untitled Document
Understanding the benchmarks
The seek time, or latency, is the length of the pause between requesting data
from the disk, and receiving the first byte. This figure may be measured in
milliseconds, but it rapidly adds up when the drive is heavily used, especially
for small fragmented files. A low latency drive will speed up access to program
and operating system files. Latency is not so important for a drive which is
only used for storing video, music, and other large files.
PC
Mark 2004 was used as a general drive performance benchmark. This benchmark
simulates tasks such as starting up Windows, copying files, and loading applications.
It gives bandwidth figures for these in megabytes per second, as well as an
overall score. The overall score provides a reliable and simple indicator of
the drives performance in most situations.
Believing the specifications
It is easy to compare the value for money of different hard drives if all you want is the biggest drive your money can buy. Unfortunately, if you are concerned about hard drive performance, there are very few ways of finding out what you are getting before you buy.
As these benchmarks of the Maxtor DiamondMax +9 IDE range show, latency tends
to increase as the disk gets bigger, but there is no correlation between disk
size and overall performance. The difference in general performance between
the 200 GB and 120 GB models is almost 20%. This means that if one drive in
a particular model range performs well, there is no guarantee that others in
the same product range will do the same.
| |
80 GB |
120 GB |
160 GB |
200 GB |
| PC
Mark 2004 |
4456 |
3663 |
4236 |
4574 |
|
XP Start Up |
6.7 MB/s |
5.9 MB/s |
6.53 MB/s |
7.13 MB/s |
|
Application Loading |
6.7 MB/s |
5.3 MB/s |
6.45 MB/s |
6.88 MB/s |
|
File Copying |
37.62 MB/s |
33.5 MB/s |
35.07 MB/s |
37.96 MB/s |
|
General Usage |
5.25 MB/s |
3.98 MB/s |
4.93 MB/s |
5.35 MB/s |
| Latency |
6.6 ms |
9.75 ms |
9.21ms |
8.71 ms |
As you can see above, the results are very different. This difference is purely
down to the amount of platters and the platter density. When buying online,
it is often impossible to discover if you are receiving a drive with say three
27GB platters or one 80GB.
Comparing the SATA 36 GB Western Digital Raptor with the 250 GB ATA100 Hitachi
Deskstar, we see that not even the rotational speed of the drive, nor the interface
standard necessarily has anything to do with performance!
| |
WD Raptor 36 GB
|
Hitachi Deskstar
250 GB |
|
Interface |
SATA 150 |
ATA 100 |
|
Rotational Speed |
10 000 rpm |
7200 rpm |
|
Cache |
8 MB |
8 MB |
| PC
Mark 2004 |
5215 |
5398 |
| XP
Start Up |
8.45 MB/s |
8.61 MB/s |
| Application
Loading |
8.97 MB/s |
8.26 MB/s |
| File
Copying |
33.55 MB/s |
40.97 MB/s |
| General
Usage |
7.68 MB/s |
6.54 MB/s |
| Latency |
5.96 ms |
6.93 ms |
To make matters worse, there is no real way of comparing benchmark results
from different drives, unless exactly the same program was used for testing,
and the same test bed system was used.
The bottom line is that manufacturers specifications don't mean a thing and
you can't compare drives by reading separate reviews. This is of course, where
a round up comes in handy.
<< Back | 1 | | 2 | | 3 | | 4 | | 5 | | 6 | | 7 | | 8 | | 9 | | 10 | | 11 | | 12 | | 13 | | 14 | | 15 |
Next >>
|
