
 |
HotHardware Test Systems |
AthlonXP All the
Way!! |
|
MSI KT3
Ultra-ARU
KT333
Motherboard
Asus AN7266-E nForce Motherboard
AMD AthlonXP
1800+.
256MB Corsair XMS2400
(2-2-5-2)
2 IBM ATA100 7200RPM
42GB HD
nVidia GeForce3 Ti500
Creative 52X CD-ROM
Standard Floppy Drive
Windows XP Professional
DirectX 8.1
Detonator 23.11 |
A Few
Words About The Benchmarks:
As weve noted in the past, weve seen a wide difference
in benchmarking scores across the net. To help
explain the scores we achieve when testing a product, we
felt it necessary to explain how we set up a system
before running the benchmarks. With the MSI KT3
Ultra-ARU motherboard, we started off by setting the
BIOS to High Performance Defaults. The memory
frequency was manually set to 266MHz. We then set
up the Two IBM Deskstar hard drives in a RAID-0
configuration with the Promise controller set for
"Performance/A-V Editing". We then installed a
fresh copy of Windows XP Professional and followed that
up with a download of all Critical Updates in Windows
Update. Next we ran Scandisk and then defragmented
the array. Windows XP has a number of video
features that make the interface more visually pleasing,
but at a cost of some performance. In our test
system we set the visual quality to "best performance"
with both the Windows Desktop Customize option and in
the Video drivers. |
 |
Benchmarking with Sisoft Sandra 2002 Pro |
Starting with
the Synthetic... |
|
SiSoft Sandra
2002 Professional is an excellent utility for comparing a
systems overall potential with comparison systems in the
suite's internal database. Below we've listed the
scores we achieved with the KT3 Ultra ARU at both default
and overclocked settings.
CPU@1.53GHz.
.jpg) |
CPU@1.70GHz.
.jpg) |
Multimedia@1.53GHz.
.jpg) |
Multimedia@1.70GHz.
.jpg) |
The first thing you might notice is how the CPU is reported
as a 1.54GHz. rather than the 1.53GHz. that the Athlon XP
1800+ is actually clocked at. This is probably a
result of the motherboard's aggressive timings, a common
occurrence in higher-end performance systems with PLLs that
have soft BIOS adjustable bus settings. In the CPU
tests, the test system easily dominated the Pentium 4 @
2GHz. with the arithmetic tests but lagged in the floating
point processes. Once overclocked, the KT3 Ultra test
system narrowed the margin, finding a balance between ALU
and FPU processing. In the multimedia based tests, the
AthlonXP 1800+ dominated, taking a commanding lead when
overclocked.
Memory@1.53GHz.
.jpg) |
Memory@1.70GHz.
.jpg) |
The memory performance of the KT3 Ultra was quite good,
yielding a sizable gain over a KT266A based system with
PC2100 DDR-RAM. When we overclocked the system to
147MHz. BUS-Speed, the RAM performance increased quite
nicely. One of the problems we've noticed though, is
that the KT333 chipset is so new that it is still identified
by Sandra as a KT266A based board. This has been noted
by the Sisoft team and currently no update is available.
With that in mind, we suspect that these scores will change
as the program receives future updates to correctly identify
the KT333 chipset.
Hard Drive - Single
.jpg) |
Hard Drive - RAID
.jpg) |
Hard drive performance with a single drive was on par with
the reference systems in the utility. The fun really
started when we tested the systems RAID performance.
Running in a RAID-0 configuration, the score leapt off the
scale, clearly surpassing anything on the on chart. We
were skeptical at first of the Promise Controller with the "Lite"
BIOS since there were no advanced settings available to
tweak the array. Obviously these scores put those
doubts to rest, showing that there was no sacrificing
performance with the Promise "Lite" BIOS.
Now that we have
a feel for how this system should perform, let's take
a look at how well it does perform.
 |
Benchmarks
and
Comparisons |
The CPU Tests |
|
Throughout this
portion of the review, you will find that we've compared
various scores of the KT3 Ultra ARU with the ASUS A7N266-E
motherboard. We like to do this to simply show how the
system performed in comparison to another test bed.
This is not done in a "shoot-out" fashion, having two
systems duke it out for motherboard supremacy but rather as
a frame of reference to show what you can expect in
performance compared to another popular system. The
first round we've run are the CPU intensive tests, that show
the systems pure CPU processing capability.
MPEG-2
Encoding Performance with Video 2000:
The first test
we ran was the Video 2000's MPEG-2 Encode test. This
is an excellent test to demonstrate pure CPU processing
performance of a motherboard since the encoding process is
completed almost entirely by the CPU itself. For this
test, the scores were compared with the ASUS A7N266-E test
system.

Although the MSI
had a slight advantage over the ASUS A7N266-E, both boards
performed well.
Next we'll focus
on OpenGL and DirectX performance and then get down to
business with the Winstones.
More
Quake 3, The Winstones & 3DMark2001 |