|
Test
System:::...
ABIT
AI7 i865 motherboard
Intel Pentium4 3.06GHz (533MHz).
512MB Corsair TwinX
ATi Radeon 9800 Pro
WindowsXP Pro + SP1 (NTFS)
Additional Hard Drives used:
Maxtor DiamondMax Plus D740X - 80GB
~ ATA-133 ~ 7200RPM - 2MB Cache
Western Digital WD800JB .........-
80GB ~ ATA-133 ~ 7200RPM - 8MB Cache
FC-Test:::...
FC-Test is a real world file copy test
created by Serguey Gromov for X-bit labs and though
far from perfect it's a reliable and repeatable test
for real world file read and copy results.
For testing I used the pattern "win.ptn"
to create files in the location "FCTest-Source"
on the drive being tested. As you can see below, this
patter creates 9006 individual files totaling 1.08GB
of data. Creating your own patterns based on files
you have on your own hard drive is simplicity itself,
but I chose one of the existing patterns so that you
can compare results at home.
In future I'll also be using the results
from writing the files to the source directory and
also running read times in addition to copy times
but I'm still evaluating that side of things for now.
The files are created unoptimised (not
sorted according to size) and are then copied to a
second folder on the same drive labeled"FCTest-Destination"
and the time taken recorded.
Each procedure is run three times with
a reboot between them to flush any cached data. Perhaps
the main complaint with this very useful little program
is that there's no control over where the data is
physically written on the platter, though having said
that when was the last time you saved data at home
and had any control over where the data was written?
Because of the possible variables we may not see pinpoint
precision in the recorded times but I think it's fair
to say the variations are no more than you'll face
in general use.
|
Western
Digital WD2500JD
|
|
Run 1
|
 |
|
Run2
|
 |
|
Run3
|
 |
|
Western
Digital WD800JB
|
|
Run 1
|
 |
|
Run2
|
 |
|
Run3
|
 |
|
Maxtor
D740X
|
|
Run 1
|
 |
|
Run2
|
 |
|
Run3
|
 |
The award for widest variations in copy
times went to the WD2500JD with a gaping 34 seconds
between best and worst. Incredibly though even the
worst time at 177 seconds was some 14 seconds faster
than the next fastest time of 191 seconds as recorded
by the WD800JB.
|
