As usual, I ran all tests 3 times and rejected the run if the results were not within 3%. I used an Athlon 1.2GHz 266MHz FSB CPU, 128MB of CAS 2.5 PC2100 running at CAS 2 (SEC and Micron), Western Digital 200BB 7200 RPM ATA/100 hard disk, D-Link 10/100 NIC and a Visiontek GeForce II GTS video card set for highest performance (vs. quality). Windows 98 SE was installed and no sound or USB was enabled. VIA 4.31, AMD 4.30 and ALi 1.74 driver (packs) were installed along with NVidia W9X_650. Windows was set for 1024×768/16bit @ 85Hz and the default disk cache was used. All tests were run back to back over a period of a few days, none were ‘left over’ from previous testing.
One comment before I start: This started out as just a review of the MSI K7T266 Pro and the VIA KT266 chipset, and to follow it as it matures. Even though I did see an improvement in performance with the revised BIOS’, I couldn’t just focus on that. Just like the KT266 Pro has improved with ‘age’ with new BIOS’ and drivers, so did the Soyo K7ADA / ALi MAGiK1 and Gigabyte GA-7DX / AMD 761 (with VIA 686B) based motherboards. It only seemed fair to see what changes they had gone through and do a new comparison. One thing about the PC business is that change is constant. I also included a VIA KT133A based Soyo K7VTA Pro for a comparison between PC133 SDRAM and PC2100 DDR SDRAM.
As a rule of thumb I use the following criteria when judging performance:
- 3% is the margin of error due to variances in the test runs
- 5% before you can really say one is really faster then another
- 10% difference is required to even start to notice any in actual usage.
So a difference of a few percentage points just doesn’t matter too much to me in the overall scope of things.
Be the first to discuss this article!