test system specs: |
|
computer hardware: |
|
processor: |
amd athlonxp 1900+ |
clock
speed: |
1.66 ghz 12.5 x 133 10 x 166 |
motherboard: |
msi kt3 ultra-aru epox 8kha+ |
chipset: |
via kt333 (msi) via kt266a (epox) |
videocard: |
MSI GeForce 3 Ti500 (MS-8850) |
Sound Card: |
N/A |
Network
Card: |
DLink DE-528CT |
Memory: |
256MB Corsair PC2400 CAS2
DDR |
Hard Drive: |
Samsung SV6004H 5400RPM UDMA100
60GB |
CDROM: |
Panasonic 48x CD-ROM CR-594-B |
Floppy: |
Panasonic 1.44MB Floppy Drive |
Heatsink: |
AVC Sunflower |
PowerSupply: |
Enermax 431Watt |
Software
Setup |
Windows 2000 Pro + SP2 VIA 4in1
4.37V DetonatorXP 22.80 DirectX 8.1 |
Benchmarks |
SysMark 2001 Business Winstone 2001 Content
Creation 2001 Winbench 99 SiSoft Sandra 2002 Pro 3DMark2001
SE Serious Sam 2 Return to Castle
Wolfenstein |
SYSmark2001 uses real world applications
concurrently as an accurate way to test the system by following patterns that
reflect the way real users work. SYSmark2001's workloads accurately represent
today's updated business usage model. The benchmark runs at a realistic
execution speed, with think times between application interactions, in order to
emulate a desktop user's interaction with the operating system and
applications.
Higher numbers denote better
performance.
Looking
at the numbers, the
MSI KT3 Ultra does get a tiny performance boost when the memory is clocked at
166 MHz with the FSB still at 133. Note that once the Epox 8KHA+ and MSI KT3 Ultra
are at the same frequency, the scores are almost identical.
Winstone 2001 |
Source: Zdnet |
|
Content Creation Winstone
2001 keeps multiple applications open at once and switches among those
applications. Content Creation Winstone 2001's activities focus on what we call
"hot spots," periods of activity that make your PC really work--the times where
you're likely to see an hourglass or a progress bar.
Business Winstone is a
system-level, application-based benchmark that measures a PC's overall
performance when running today's Windows-based 32-bit applications on Windows 98
ME, Windows NT 4.0 (SP6 or later), Windows 2000, Windows Me, or Windows XP which
runs real applications through a series of scripted activities and uses the time
a PC takes to complete those activities to produce its performance
scores.
As we can see here, Content
Creation 2001 and Business Winstone 2001 verify what SysMark2001 says. The KT333
does give a very slight performance boost in office based applications when the
memory is set to run at 333 MHz. When the two boards are clocked at the same speed,
they perform identical to each other.
WinBench 99 is a
subsystem-level benchmark that measures the performance of a PC's graphics,
disk, and video subsystems in a Windows environment. WinBench 99's tests can run
on Windows 95, Windows 98, Windows NT, Windows 2000, and Windows Me
systems.
WinBench 99 v1.2 Benchmark Results |
|
MSI KT3 Ultra-ARU |
Score |
Ranking |
1. |
Business Disk - 133/266 MHz |
5750 |
|
1a. |
Business Disk - 133/333 MHz |
6010 |
|
1b. |
Business Disk - 166/333 MHz |
5850 |
|
|
|
|
|
2. |
High-End Disk - 133/266 MHz |
15800 |
|
2a. |
High-End Disk - 133/333 MHz |
16000 |
|
2b. |
High-End Disk - 166/333 MHz |
14300 |
|
|
|
|
|
3. |
Business Graphics - 133/266 MHz |
835 |
|
3a. |
Business Graphics - 133/333 MHz |
841 |
|
3b. |
Business Graphics - 166/333 MHz |
876 |
|
|
|
|
|
4. |
High-End Graphics - 133/266 MHz |
1590 |
|
4a. |
High-End Graphics - 133/333 MHz |
1600 |
|
4b. |
High-End Graphics - 166/333 MHz |
1660 |
|
With the memory at 333 MHz, the HDD portion of this benchmark
does get a very little performance increase, however as we see when overclocked, the
scores are lower. This is likely due to the fact that HDD's
don't work as well with higher PCI timings.
For AGP/graphics performance, the more bandwidth
the better. When the FSB is in sync with the memory, the AGP seems to love
running higher then usual.
Sisoft Sandra 2001 |
Source: Sandra |
|
Sandra is designed to
test the theoretical power of a complete system and individual components. The
numbers taken though are again, purely theoretical and may not represent real
world performance.
SiSoft Sandra 2001 Benchmark Results |
|
MSI KT3 Ultra-ARU |
Score |
|
Multimedia Benchmark
(FSB/Memory) |
|
1. |
Integer SSE - 133/266 MHz |
9055 it/s |
1a. |
Integer SSE - 133/333 MHz |
9055 it/s |
1b. |
Integer SSE - 166/333 MHz |
9055 it/s |
2. |
Floating-Point SSE - 133/266 MHz |
10657 it/s |
2a. |
Floating-Point SSE - 133/333 MHz |
10657 it/s |
2b. |
Floating-Point SSE - 166/333 MHz |
10663 it/s |
|
CPU Benchmark (FSB/Memory) |
|
3. |
Dhrystone ALU - 133/266 MHz |
4578 MIPS |
3a. |
Dhrystone ALU - 133/333 MHz |
4608 MIPS |
3b. |
Dhrystone ALU - 166/333 MHz |
4601 MIPs |
4. |
Whetstone FPU - 133/266 MHz |
2309 MFLOPS |
4a. |
Whetstone FPU - 133/333 MHz |
2309 MFLOPS |
4b. |
Whetstone FPU - 166/333 MHz |
2311 MFLOPS |
|
Memory Benchmark (FSB/Memory) |
|
5. |
Integer ALU - 133/266 MHz |
2066 MB/s |
5a. |
Integer ALU - 133/333 MHz |
2087 MB/s |
5b. |
Integer ALU - 166/333 MHz |
2530 MB/s |
6. |
Float FPU - 133/266 |
1977 MB/s |
6a. |
Float FPU - 133/333 MHz |
2004 MB/s |
6b. |
Float FPU - 166/333 MHz |
2379 MB/s |
Since the CPU speed is the same 1.66 GHz, the almost identical scores
in the Multimedia and CPU benchmarks isn't surprising. As for the memory bandwidth it seems that the CPU/FSB cannot take advantage of the
bandwidth available. At 333 MHz, the numbers are very close to that of the
266 MHz. With the FSB running at 166, the bandwidth increases quite a
bit!