Stability

After plugging in the motherboard and doing a fresh install of Windows Millennium and the K7VTA drivers, we were met with a flurry of lockups. A quick call to Soyo produced the VIA 4.04 AGP drivers, which immediately fixed our stability problems. In our testing we have found that VIA KT133 based boards are generally not as stable as other chipsets. They are very sensitive to installed hardware and, overall, have been less stable in our tests. The K7VTA breaks out of this mold. It is one of the more stable boards we have tested and easily the best KT133 based board. Seventy two hours of RC5 running in the background and looping 3D Mark 2000 resulted in only two crashs. Overall stability is quite good and we experienced few hardware related problems.

 

Overclocking:

Ever since the release of the original SlotA Athlons, AMD CPU's have been the choice of overclockers. Tthe simple reason is that they are not multiplier locked. As we have seen with other KT133 based boards the EV6 bus and KT133 chipset are very sensitive to overclocking, leaving the ability to change the multiplier as the only means for bumping up the megahertz. Offering multipliers of 5 through 12.5X combined with the ability to change the CPU core voltage from -0.1v through +0.250v in increments of 0.025v is an excellent combination for great overclocking abilities. In order to stop remarking of its' chips, AMD has begun cutting the L1 bridges on the surface of the ceramic package. In order to re-enable multiplier adjustments, all that has to be done is to reconnect the L1 bridges, as shown below.

 

In order to re-enable multiplier adjustments on AMD CPU's, the L1 bridges must be reconnected. To do this, simply find a #2 HD pencil with a fine point and draw a line over each set of L1 bridges to reconnect them. This re-enables the motherboard's ability to change the CPU multiplier.

In testing we were able to crank up our Duron 700 to 950 before stability was compromised, and the motherboard was quite ready to go farther. The excellent stability of the board compliments its' good overclocking options, producing a board which has above average overclocking abilities. The only hiccup we experienced while overclocking it was lockups in CounterStrike. The system refused to run CS in an overclocked state (even 750Mhz), but this instability was not reproducible in any other application.

The placement of the multiplier changing DIP switches is our only overclocking complaint about this board. Placed above the IDE connectors they are nearly impossible to reach with the board installed. This is ever harder to accept when other makers are now allowing multiplier adjustments in the BIOS.

Conclusion