Intel Pentium 4 2.6c 800MHz FSB - More Benchmarking

(Page 3 of 4 )

(Benchmarks Continued...)

PiFast Pi Calculation Benchmark

This is a benchmark that I really like to use, especially for CPU reviews. It is as simple as it's title insinuates. It measures the time necessary for the system to calculate Pi. This of course is excessively CPU dependant, but we've found the system bus to also play a role in the score.

Here's the settings we used:

Program : PiFast version 4.2, by Xavier Gourdon
Computation of 10000000 digits of Pi
Method used : Chudnovsky
Size of FFT : 1024 K
Physical memory used : ~ 61355 K
Disk memory used : ~ 0.00 Meg

Ok, let's take a look at the results:

Now here we notice almost no difference at all in the overclocked scores of the two CPU's.  This can be attributed to the fact that calculations are dependant solely on the operating frequency of the CPU to carry out its calculations.

PCMark2002

PCMark2002 is an entire system benchmark that pays special attention to CPU and memory performance. Because we're reviewing a CPU and the bus of the CPU plays a major role in memory performance we felt it appropriate to include this benchmark. This is a relatively older benchmark but it's scaled well with the times.

Here again we can see the advantage of the higher bandwidth showing its performance edge.  Not a substantial increase, but a noticeable one.

Unreal Tournament 2003 Demo

I can already hear your question. "This is a 3D game so why is it a valid benchmark for a CPU?" Well, for two reasons. First, we're benchmarking it in a system where the only difference is the CPU, meaning that any differences you see in results are directly attributed to the CPU. Second, we're using the HardOCP.com UT03 benchmarking utility's "CPU" setting at a resolution of 640x480, basically ruling out our R9700 video card as a bottleneck and making plenty of room for raw CPU performance.

Same story different benchmark.  The HT does not have an impact at all really noticing the identical average frame rate by both the 2.6 with and without HT enabled.  Also, we again see that the extra front side bus offers a decent gain.  With this test running at only 640x480 with the lowest quality settings it's safe to say that the CPU performance is ruling the results.

Aquamark

Aquamark is a benchmark based on the engine behind the Aquanox game. We've run the default benchmark utilizing no FSAA and a resolution of 1024x768. What we look to gain here is basically the same as our reasons for running UT03, well, that and it just looks pretty. 

Now with these results, taken using 1024x768 default settings, the extra FSB doesn't appear to have helped at all.  That could be for two reasons A, the synthetic benchmark (all be it based on a real world gaming engine) is unable to utilize the extra available bandwidth, or, and this is my guess, we're seeing the video card be the bottle neck.  I find that remotely hard to believe as this is a level III OCS Radeon9700, but it's certainly appearing that way.

Next: Conclusion >>
 

More Computer Processors Articles
More By Jim Miller

Comments On: Intel Pentium 4 2.6c 800MHz FSB

>>> Be the FIRST to comment on this article!