|
TESTING METHODOLOGY:
We have seen quite a
variation in benchmark scores from one site to the
next, so we feel it is necessary to explain
exactly how we configure each test system before
we run any benchmarks. The first thing we
did was enter the system BIOS and set the board
to it's "High Performance Default" settings. We
then set the Memory clock to "+33MHz", and set the CAS Latency
and memory timings to 2-2-5-2,
with Bank Interleaving set to "Auto" and a 1T Command
rate. The hard drive
was then formatted, and Windows XP
Professional was installed. After XP was
completely installed, we hit the Windows Update
site and downloaded all of the available
updates, with the
exception of Windows Messenger.
Then we installed all of the necessary drivers, disabled Windows Messenger, disabled
Auto-Updates, set a 768MB swapfile and disabled System Restore.
Lastly we set the Visual Effects to "best
performance", installed the benchmarking software,
defragged the hard drive and ran the tests at the
CPU's default and overclocked speeds.
 |
|
The Hot Hardware Test Systems |
|
It took 30
years to hit 1GHz, and 18 months to
hit 2GHz! |
|
|
HARDWARE USED:
VIA P4X266A
P4XB
Intel S478
Pentium 4 1.9GHz. (1900MHz.)
256MB Crucial
PC2100 (CAS 2, 1T) @ 133MHz.
GeForce 3
Ti500 (21.85 Drivers)
On-Board Sound
3Com NIC
IBM 7200RPM
30GB HD
Creative Labs
52X CD-Rom
Standard Floppy
Drive
Windows XP Pro
(With all current updates)
DirectX 8.1
VIA 4-in-1's v4.37
VIA AGP Update
4.10
|
 |
|
Performance Comparisons |
|
Time for
some numbers... |
|
|
The first benchmarks
we ran were using the very popular subsystem tests
that are part of SiSoftware's Sandra Benchmark
suite. We also overclocked our 1.9GHz. CPU
(19x100) to 2.08GHz. (19x108) and ran a second set
of numbers for the sake of comparison.
SiSOFT SANDRA:
CPU @ 1.90GHZ.
CPU @ 2.08GHZ.
In Sandra's CPU tests,
out 1.9GHz. P4 sped along nicely, performing in
line with the listed reference systems.
When we overclocked our CPU however, we pulled
ahead of the pack, albeit by a small margin.
M.M. @ 1.90GHZ.
M.M. @ 2.08GHZ.
We see more of the
same in the Multimedia tests. VIA's
unlicensed chipset seems to have no trouble
letting the P4 spread it's wings.
Again, while overclocked over the 2GHz. mark, we
outpaced all of the reference systems. (Keep in
mind that the version of Sandra we used did not
have any Northwood or high-end Athlon XP scores
listed.)
Before we move onto
the very important memory bandwidth performance,
we figured we'd throw in a hard drive benchmark
for good measure. The VIA VT8233 Southbridge
handled our IBM hard drive well, as you can see,
it's performance was in-line with similar drives
in it's class.
MEMORY @ 1.90GHZ.
MEMORY @ 2.08GHZ.
I'm sure many of you
curious gurus out there jumped straight to these
scores! For a long time, P4 owners were
forced to pay a price premium on RDRAM if they
wanted a top-performing system...well, not
anymore. While the RDRAM based reference
system delivered higher bandwidth, you must keep
in mind that DDR SDRAM has lower latency.
This means that in some instances where memory
bandwidth isn't being taxed, a P4 with DDR SDRAM
may be the quicker system. Having lower
latency somewhat offsets the bandwidth difference...
 |
|
Performance Comparisons |
|
Time for
some numbers... |
|
Next up, we wanted to
see how the P4XB would handle a little MPEG
encoding. We ran the MPEG Encoding test in
MadOnion's Video 2000 benchmark. From this
point forward, we'll be comparing the VIA P4XB's
performance to a similarly configured system that
used an Intel i845 DDR chipset.
VIDEO 2000 MPEG ENCODE:
In the encoding test,
the VIA board just barely outperformed the i845.
The performance difference is well within the
margin of error for this test though. Either
chipset would be fine if Video editing / encoding
is your gig.
Let's move onto another of MadOnion's benchmarks
now,
the very popular 3D Mark 2001...
3D
MARK 2001:
We ran 3D Mark 2001
through a default run (1024x768x32 NO AA), and saw
the VIA P4XB outpace the i845 by about 5%.
This is fairly significant considering that with
the exception of the motherboard, all of the other
hardware used was identical.
|
|
More Gaming
and the Stones...
 |
|