|
It was time to get a little more
serious about our testing and fire up some real world
applications and benchmarks, that would stress our new found
clock speed and showcase the performance advantages that the
Vapochill could bring to the table.
 |
|
Benchmarks -
PCMark 2002, 3DMark 2001SE and Comanche 4 |
|
Percentage Gains In
Performance |
|
MadOnion's PCMark 2002
Processor and Memory Performance tests were used to look at
the relative performance levels of the 3.3GHz Vapochilled P4
versus its stock speed counterpart. The CPU test in
particular utilizes the following benchmark components.
CPU Test:
Here we see a
modest 14% performance gain in both the Memory and CPU
Tests. This test suite certainly has a synthetic feel
to it. Regardless, the scores are a indicative of real
world performance metrics for things like JPEG compression
and MP3 Audio conversion.
In the
3DMark2001SE test, we also overclocked our Radeon 9700 Pro to
a core clock of 375MHz and memory speed of 700MHz DDR, just
to help alleviate any bottleneck the graphics pipeline
could be imposing on this test. However, we wanted to
run the stock benchmark, as many folks do, who download the
demo version of 3DMark, so you could see things from a well
known reference point. All told, we're looking at a
10% gain. However, keep in mind you can't
possibly score an "absolute zero" 3DMark in this test, so
the impact is greater than it may seem at first glance.
Novalogic's
Comanche 4 is far more demanding on the host processor of a
given system, rather than the graphics engine. In
short, the game is very much a CPU hog. There are a
lot of physics calculations that are made during in game
action, as well as calculating various levels of enemy AI.
It really beats on the CPU. As a result, we see a more
significant impact with the Vapochilled Pentium 4 3.3GHz
setup, which is approximately 18 - 19% faster.
UT2003, Quake 3 Benchmarks And The Rating
| 
