The BIOS:
The PF21 Extreme uses the Phoenix Award Bios and
includes a pretty wide range of tweakability so most should
be quite happy with the overclocking options provided. Below
is the default BIOs screen, we'll move through some of the features
as we go along.
Most enthusiasts are going to go straight to the
CPU frequency and voltage control, so that is where we will
start. There is a pretty wide range of adjustment here for the
CPU you have voltage options ranging from 1.125v to 1.5875v.
PCI-E x16 slot gets options ranging from 1.5v to 1.675v, and
for the DDR2 your provided options between 1.8v up to 2.2v.
As you can see, there are a lot of options for voltage control,
and this should allow for some fine tuning to find the most
stable overclock. you also have options for the PCI clock and
CPU clock. The CPU FSB can be adjusted in 1Mhz increments from
200 up to 510MHz, this gives you a HUGE range to play with and
maximize your CPU's OC.
In the advanced chipset features is where you
will find your memory settings, you have full control over memory
timing which will allow you to fine tune your memory in your
quest for a stable OC. Due to the chipset used, the board only
supports 400MHz or 533MHz memory, but again this is due to the
chipset used. I originally tried setting this board up with
some kingston HyperX 5400 and the board simply would not boot
with the 675MHz RAM installed. Once I installed the 533MHz Patriot,
it booted right up.
The rest of the BIOS is pretty standard, so I
wont be going into detail about it. You have your standard options
that we are all familiar with and they don't provide anything
out of the ordinary.
Test System:
The system build for this review is, ECS PF21
Extreme motherboard, Intel P4 550 3.4GHz, 1GB Patriot DDR2,
HIS X800XL, 2 x WD 74GB raptor HDD, Lite On DVD burner, Lite
ON DVD ROM, Cooler Master RealPower 550 PSU.
Overclocking:
The very first thing I tested was the overclocking
ability of this board. I will admit, I am not an avid overclocker
and all of my previous overclocking exp. is with AMD processors.
So I did not do a LOT of tweaking to determine a max stable
overclock. The default FSB and multi for this processor is 200
x 17 which is 3.4GHz. In my overclocking test I made adjustments
to the FSB only and was able to get a rock solid 4GHz with 235
FSB x 17 multi. At 236 FSB windows would start and run but eventually
crash. With some fine tuning and voltage tweaking, I'm sure
this board would go well beyond 4GHz with this processor, but
not being an experienced Intel overclocker, and this being my
only LGA chip, I didn't want to risk killing it or the board.
Below are screen shots of the default speed and the stable overclock
I achieved.
For a novice overclocker a 600MHz increase in
speed is nothing to sneeze at, an experienced overclocker could
probably get a very impressive OC out of this board.
Benchmarks and system testing:
All benchmarking and system testing was done at
default setting, no overclocking was done for any of the benchmarks.
Software used for testing will include.
SiSoft Sandra
2005 - We will be testing the CPU, MMX, and memory speeds,
using the 32-bit 2005 version.
Business
Winstone 2004- The ZD Winstone suite is a script that runs
a series of actions and calculates a final score that measures
a PC's overall performance. Higher numbers are better.
PiFast
- We used a computation of 10000000 digits of Pi, Chudnovsky
method, 1024 K FFT, and no disk memory. Note that lower scores
are better, and times are in seconds. Note that lower scores
are better, and times are in seconds.
CDex 1.51
- Blue October- Argue With a Tree CD1 was ripped into one .wav
file. We then encoded that .wav file into a 320Kb/s sample rate
MP3. Times are in minutes:seconds, and lower is better.
Doom 3 , FarCry, UT2K4 @ 640x480, LQ Settings
- While higher resolutions tax the video card, lower resolutions
rely on CPU and subsystem speed. These results are real-world,
and higher scores are better. Bench'emAll
was used to collect numbers.
All benchmarks will be run a total of three times
with the average scores being displayed. For comparison of similar
systems take a look at some of our other benchmarks that can
be found here.
Sisoft Sandra 2005
The Sisoft benchmarks are pretty consistent with
other boards we have tested using this chipset as seen here
. The Albatron board does do a bit better in the memory bandwidth
test, but it was tested with faster Kingston memory.
Business Winstone 2004
This test was ran three times and the scores averaged,
the higher the score the better the system performs in these
tests.
Again,
this score is on par with other 925x boards we have looked at.
PiFast
| |
Time,
minutes:seconds |
| ECS
PF21 |
48.34 |
For math calculations this board and cpu combo
scores very well and is right on par with other more well known
boards we have looked at.
