|
|
Abit KX7-333/KX7-333R DDR Athlon
Motherboard
Last updated: 10/3/2002
The Western Digital WD800 drive has 8 MBytes of cache'
compared to the other drives with 2 or less MBytes of cache. I have
no explanation for the last Cached Disk benchmark except roughly the same
results were obtained in repeated tests and that WinTune is measuring the
Windows 98 disk caching. My impression of performance is that uncached
disk test on the 80 gig WD drive is more in-line with what a user would perceive
as compared to the other drives, but not as extreme as the numbers would
indicate--it flies. The problem is in Wintune 98, an old benchmark. Let's
move on to more tests with newer benchmark software from Zeff-Davis.
| Config |
1 |
2 |
3 |
4 |
5 |
6 |
| Mother-board |
EpoX
8KHA+ KT266A |
Abit
KX7-333
KT333 |
| Operating
System |
Win
XP |
Win
XP |
Win
XP |
Win
XP |
Win
XP |
Win
XP |
| Hard
Disk |
WD300BB
30 GByte
7,200 RPM
ATA/100 |
Maxtor
53073H6
30 GByte
7,200 RPM
ATA/100 |
WD800
80 GByte
7,200 RPM
ATA/100 |
WD800
80 GByte
7,200 RPM
ATA/100 |
WD800
80 GByte
7,200 RPM
ATA/100 |
WD800
80 GByte
7,200 RPM
ATA/100 |
| Hard
Disk Config. |
Single
Drive
Normal IDE
Port |
Single
Drive
Normal IDE
Port |
Single
Drive
Normal IDE
Port |
Single
Drive
RAID IDE
Port |
Dual
Drives
RAID 0 Striped |
Dual
Drives
RAID 1
Mirrored |
| Business
Winstone 2001 |
50.4 |
49.5 |
53.9 |
57.8 |
59.7 |
57.6 |
| Business
Disk WinMark 99 |
4277 |
5350 |
8748 |
12176 |
17680 |
9360 |
| High-End
Disk WinMark 99 |
18883 |
16220 |
26480 |
30460 |
36040 |
24550 |



The Winstone tests show that the KX7-333 with it's KT333 chipset and faster
processor is neck and neck with it's predecessor, the EpoX 8KHA+ and the KT266A
chipset. Any real difference in the benchmarks is caused by the much
faster WD drive. Surprisingly, the Highpoint RAID IDE interface is faster
than the normal one managed by the VIA Southbridge. This may be a driver,
chipset, or BIOS shortcoming/issue. As expected, the highest performance
is obtained by putting two of the WD 80's on the RAID interface in a RAID
0 configuration and a performance penalty is paid when mirroring the drives.
For my use, the performance gained with one of the WD drives on the normal
IDE interface was very noticeable and is more than sufficient--"LIGHTNING
FAST" is what it says on the box the retail version of the drive comes
in and that is the way I would characterize it. The disadvantages of
the additional cost of a second drive, the reduced reliability that a RAID
0 configuration entails--if one drive fails, all data is lost, which makes
the combination twice as likely to fail over a given time period as compared
to a single drive--outweighs the advantages of the increase performance of
a RAID 0 configuration. Besides, in human terms, the drive is so fast
on the normal IDE interface it is hard to perceive the difference when two
of them are running in a RAID 0 configuration--very fast is very fast. Furthermore,
the inconvenience of having to wait for the Highpoint BIOS to initialize and
detect the drive--that's why I don't like SCSI drives, which boot in a similar
manner--during boot-up outweighs any speed advantage of putting a singe drive
on the RAID interface. The RAID capability is useful for mirroring drives
on a file server/Unix box.
One usually sees page after boring page of benchmarks in
many motherboard reviews. Often these benchmarks do not provide a clear
picture of what one can actually expect. Many of them are so focused
on competing products that they do not compare motherboards and systems that
are further separated in technology and time. In many cases, the real
differences in perceived performance, despite the benchmarks numbers, are
actually so small that they are almost insignificant to most computer users.
Benchmarks mean different things to different people depending
on what they do with a computer, their experience, and how they interpret
the numbers. If you do a lot of number crunching and graphics for programming,
engineering, science, etc. or are addicted to games, any reasonable increase
of speed is probably significant.
The Wintune 98 benchmarks are a step in the right
direction, but even these numbers do not completely depict what one will
perceive as a noticeable speed difference while sitting in front of a computer
with this motherboard as compared to competing products and others from generations
past. Perceived performance does not have a linear relationship to
the results of benchmarks tests. Benchmarks measure computers, not
human beings.
If you are an average business/home computer user, you
would probably not see any noticeable difference in speed between this motherboard
and quality DDR motherboards using the most recent Athlon chipsets from competing
manufacturers. Any speed difference between computer #4 in the Wintune
benchmarks and this one (#5) with a comparable drives are almost insignificant
to the business user. On the other hand, if you have an aging computer
with 500 MHz K-6 processor, 64 MBytes of memory, and a 5,400 RPM hard disk
drive, or something like that (sell it now or forget selling it), or a computer
like computer #2 to a somewhat lesser extent, a computer with this generation
of motherboard, processor, memory, and, especially, the WD 80 gig hard disk
drive will make the average person think that maybe it's finally time for
a major upgrade or a new computer? If you want to really soup-up a
more recent computer get a hard disk with larger disk cache.'
It would be wise to see a comparable computer for yourself
and put your hands on it for awhile before reading too much into any these
numbers (and others) and descriptions before deciding to upgrade.
< Previous | Contents | Top | Next
- Stability >
|
|