Test System
Specs: |
Computer: |
AthlonXP 1800+ (1.53 GHz at 1.7)
256MB OCZ PC2400
DDR Ram
EPoX 8KHA+
MSI StarForce822
(GeForce3 200/460)
SB Live! Platinum 5.1
Maxtor
40GB DiamonMax+ |
Software: |
Windows 98SE
Via 4in1 4.35V
DetonatorXP 22.80 |
Prcessor Settings: |
11.5x133 MHz = 1533 MHz
12.5x133 MHz = 1666 MHz
10.0x150 MHz = 1500 MHz
11.5x150 MHz = 1725 MHz |
Benchmarks: |
SiSoft Sandra 001
3DMark2000 Ver 1.1
3DMark2001
Quake III Arena Ver 1.17
Return to Castle
Wolfenstein |
I'm limited to a 12.5 multiplier bor benchmarking but was
able to overclock the FSB to reach 1725 MHz.
SiSoft Sandra
Sisoft Sandra: CPU Benchmark |
Motherboard: |
Setting: |
ALU: |
FPU: |
Epox 8KHA+ |
11.5x133=1533MHz |
4244 MIPS |
2130 MFLOPS |
Epox 8KHA+ |
12.5x133=1666MHz |
4612 MIPS |
2315 MFLOPS |
Epox 8KHA+ |
10.0x150=1500MHz |
4163 MIPS |
2090 MFLOPS |
Epox 8KHA+ |
11.5x150=1725 MHz |
4789 MIPS |
2400 MFLOPS |
Sisoft Sandra: Multimedia Benchmark |
Motherboard: |
Setting: |
Integer: |
Floating point: |
Epox 8KHA+ |
11.5x133=1533MHz |
|
|
Epox 8KHA+ |
12.5x133=1666MHz |
|
|
Epox 8KHA+ |
10.0x150=1500MHz |
|
|
Epox 8KHA+ |
11.5x150=1725 MHz |
|
|
Sisoft Sandra: Memory Bandwidth Benchmark |
Motherboard: |
Setting: |
Int Buffered: |
Float Buffered: |
Epox 8KHA+ |
11.5x133=1533MHz |
|
|
Epox 8KHA+ |
12.5x133=1666MHz |
|
|
Epox 8KHA+ |
10.0x150=1500MHz |
|
|
Epox 8KHA+ |
11.5x150=1725 MHz |
|
|
As we can see, the AthlonXP1800+ scales very
well even at stock, unfortunately the Mutli-Media bench is pretty much
useless... Still, it does put the Athlon in a good light. In all honesty, I
don't really understand how this new version of Sandra works. From now on, I think I'll be downgrading to the older version
2001TE. Take this how you will. Let's take a look at some 3D benchmarks.
3DMark2000
MadOnion 3DMark 2000 Ver 1.1 |
Setting: |
3D Marks |
11.5x133=1533 MHz |
11037 |
12.5x133=1666 MHz |
11271 |
10.0x150=1500 MHz |
11127 |
11.5x150=1725 MHz |
11403 |
3DMark2000 doesn't really seem to see the differences between
the systems. Still, there is a 400 point difference between the slowest and
fastest systems which is more than the margin of error.
3DMark2001
MadOnion 3DMark 2001 |
Setting: |
3D Marks |
11.5x133=1533 MHz |
7700 |
12.5x133=1666 MHz |
7866 |
10.0x150=1500 MHz |
7774 |
11.5x150=1725 MHz |
7993 |
3DMark2001 can differentiate between the different
configurations... Still, the scores are a lot closer then I thought they would be. Maybe
even with a GeForce3, I'm videocard limited?
Let's see what Quake III
thinks...
Quake III Arena Fastest Setting |
Setting: |
demo001 |
nv15demo |
11.5x133= 1533 MHz
|
265.6 fps
|
76 fps
|
12.5x133= 1666 MHz
|
272.5 fps
|
79.3 fps
|
10.0x150= 1500 MHz
|
268.6 fps
|
77.3 fps
|
11.5x150= 1725 MHz
|
282.7 fps
|
81.5 fps
|
Yes I knew good ole Quake III wouldn't let me down. As we can see here,
there is quite a gap between the slowest and the fastest settings. Next up is the
very popular Return to Castle Wolfenstein
based on the Quake III engine.
Return to Castle Wolfenstein Max 640x480 |
Setting: |
|
|
11.5x133= 1533 MHz
|
|
|
12.5x133= 1666 MHz
|
|
|
10.0x150= 1500 MHz
|
|
|
11.5x150= 1725 MHz
|
|
|
RTCW confirms everything else we've seen here today. The
XP CPU does scale quite well. Well what else really needs to be said about the
AthlonXP processor?
Even at stock speeds, it is a force to be reckoned with and when it's overclocked nothing can compete with it! It's the fastest processor out
there to date. Ever since AMD released the Athlon line of
processors, they've been setting the tone. Sure things might change with the release
of the Northwood Pentium 4 with it's higher bus speed and 512KB L2
cache, but then again AMD will counter with it's Thoroughbred. It's certainly a great
time to be a consumer right now!