KT266A – New or Just Improved?
When the first VIA KT266 Socket A DDR SDRAM based mainboards hit the market, there seemed to be quite a bit of disappointment expressed by a number review sites and users in regards its overall performance. MSI, who was one of the first to produce a KT266 based mainboard, reacted quickly and released “performance” beta BIOS’s for users to try. Then a final production version with those enhancements was released, all in the effort of getting a bit more performance from the mainboard and chipset. MSI even went as far as posting an article on their site on how to “tweak” the BIOS for the best performance (along with warnings that your system may not be stable). But when you look at the overall performance of the VIA KT266 chipset, it really was right in line with the AMD 760 and ALi MAGiK 1 chipsets. They are all so close that the end user would never really even notice the difference while using one system or another. But speed seems to be what sells both systems and web site advertising.
So what is the big deal if they are really all so close that you would never really notice which one is a bit faster or slower? My opinion is that it is all marketing hype, both on the part of the manufacturers and the review sites. We seem to be an industry driven by performance, where a couple points or Frames Per Second in benchmarks will brand one product as a winner and another as either an also ran or a loser. But think about it. When those results are within only a few percent of each other, are they really going to seem any different to the person actually using the system? I think not. But since we do seem to be a performance driven industry, and that is the “glitz” that may bring a reader to a site, reviewers will be prone to use it as “bait”. The same goes for the manufacturers. If that is what they feel draws the buyer to their product, then they will also promote performance, both in their own press releases and also in choosing which sites are sent review products. It’s all about money – review sites are paid by page hits and manufactures earn their money by selling more product. Well, maybe I’m being a bit harsh, since not all manufactures or sites behave that way, and we are even seeing a trend of offering or promoting more or different features rather than just performance.
Early on there were rumors of a revised VIA KT266 chipset, some were saying it would be called the KT266A while others were saying it would be just a minor stepping level change. Even in the industry there was confusion. I can remember one day being told that VIA would release a KT266A chipset in the morning and then hours later being told that VIA Taiwan had denied it. Then one day out of the blue VIA not only announced the “new” KT266A but a number of sites had been supplied with test mainboards to review, and had posted testing results. Now, I looked hard at those results and found them not only to be widely varied but also what I would consider flawed. Most seemed to focus on not just comparing the KT266 to the KT266A, but were primarily concerned with how the KT266A compared to the other Socket A chipsets. There were also a number of reviews that focused more on memory performance than overall ‘real world’ application performance. Since a system is a sum of its components, and users judge performance by how fast applications run, I just don’t feel that judging on memory performance alone is all that important. Shouldn’t the primary concern be the difference between the KT266 and KT266A, and how it affects the actual real world performance? Are we not initially interested in what gain, if any, can be expected from the new chipset? After that has been settled, then it might be a good idea to see how the new chipset compares to the others, but not before.
And what about the test systems used? If we truly want to know specifically what the difference is between the two chipsets, would it not make sense to compare the SAME mainboard with both chipsets? Otherwise you might be comparing apples and oranges, since you would not be able to tell whether any change in performance is due to the difference in mainboard and BIOS design, or the chipset alone. Furthermore, the KT266A mainboards supplied were VIA reference designs, so there was no way of knowing how they would relate to actual production parts. One other interesting thing to note is that VIA calls it a new chipset, and it is shown that way in VIA’s press releases and their own pictures. However, if you happened to look at any pictures published by the review sites it was still marked as a KT266 (printing on the chip), and just had a new stepping level marking of CE instead of CD. So is it really a “new” chipset with superior performance, or just a simple revision with a minor gain in performance?
Be the first to discuss this article!