Comparison testing would not be on equal footing, as there is only 1 Pentium 4 in house, so results will simply be posted without comparison to unequal platforms. You can compare these results with your current rig to see how this board performs. The only comparison shown details the performance difference between a unclocked system (currently at 2.53GHz) and an overclocked system (the overclocked core is 2.79GHz). Your mileage may vary when recreating these tests, depending on your RAM and components.
Test bed
- AOpen AX4PE Max
- 2.53GHz Pentium 4
- 1024MB DDR333
- AOpen Geforce4 Ti4200
- 120GB Western Digital 1200JB
- Soundblaster Audigy
We took a screenshot of WCPUID to grab the current unaltered settings of the FSB. We are currently running a quad-pumped FSB of 133MHz x 4 = 533MHz. The system multiplier is 19X (133MHz x 19 = 2527MHz).
We used Futuremark’s 3dMark2001 to test gaming performance, as it is rather standard. An AOpen GeForce4 Ti4200 was used without any overclocking or registry hacks, and was using nVidia’s 41.09 drivers. AGP aperture was set to 128MB in the BIOS, and AGP fast writes was not enabled. When recreating this test, your mileage may vary.
SiSoft’s Sandra is a useful utility for comparing motherboard performance, and we used it to measure the memory bandwidth’s increase when overclocking the FSB. There are many parameters regarding memory performance; the most general being access time and memory bandwidth. An upcoming article will be comparing detailed memory performance, but for the purposes of this comparison all memory is running at CL2.5. Below is the difference using the same RAM between the default FSB (133MHz) and the maxumum stable FSB with this memory (146MHz).
Sandra was used once again, but this time to test the drive performance of a Western Digital 1200JB hard drive. This drive has an 8MB cache, and was used on the ATA133 header.
HDTach is a more accurate drive benchmark, in my opinion. It displays a graph indicating throughput as it relates to the location on the disk. Since the disk is spinning at a constant rate, the bits on the outside of the platter will pass the head more quickly than the inside. This may be more of a benchmark of the drive rather than the motherboard, but it demonstrates that the drive controller is not a bottleneck in performance. Listed here is a Western Digital 1200JB and an IBM 60GB Deskstar. Again, your mileage may vary, especially when using different hard drives.
