VIA KT266 comes under fire

Often on the Internet, special favour is given to websites that can turn a review around in time for a product release. For example, companies like nVidia and AMD often let various hardware gurus like Dr. Pabst and Anand in on new releases and give them ample time to prepare coverage for the morning of the announcement. This approach is favourable not only to the websites in question but also to the consumer, who can tell right off the bat whether Product X is good or not. At least, that's the theory. In actual fact, people who rush out and buy Product X on D-Day may feel more like a guinea pig than a consumer before long. The most notable embarrassment in recent times was Intel's flustered notrelease of the Pentium III 1.13GHz processor, which was eventually withdrawn after issues arose that pegged not only its performance, but also its stability. These were brought to light not by Intel QA but by the hardware community that tested it. The latest fly in the ointment is the VIA KT266 motherboard chipset. The theory worked as far as hardware websites receiving and rating the boards highly. The consumer, on the other hand, discovered that the platform is rife with instability and questionable performance. One of the consumers just so happened to be OCWorkBench, a respected enthusiast website. According to the author of its latest update, there are but a few KT266 boards actually on the market, and "most of them are plagued with performance and stability issues". That's fairly damning. Also, manufacturers are apparently "ironing out the problems", despite having released the boards. Call us cynics, but we reckon it's quite telling that VIA are now stuck in the middle of two possible chipset blunders, which are ostensibly the result of poor QA. It was only a few days ago that an issue with the Southbridge of KT133A boards was uncovered, yet again by enthusiasts - an issue that could cause corrupted data in large files. However, the blame might not lie with VIA. OCWorkBench, again. "Over at our labs here in Singapore and Taipei, we have some early retail boards. Preliminary results are VERY unacceptable for a DDR board and we have decided to postpone the reviews. From our usual sources of information, a R&D engineer has mentioned to me that a number of brands rushed out their boards. Only after a week or so, there is an ECN issued by VIA regarding a change of registers - that is what governs the performance issues. (this happened during my tests of the first and second Gigabyte 7VTX board, as I was told by Gigabyte that they received a new reference BIOS from VIA). It looks like everyone is now rushing out there to correct these problems." There's more at OCWorkBench. The lesson here perhaps is to let someone else be the guinea pig for a while. Motherboards are pretty complicated creatures, and need a few weeks to make sure all their cylinders are firing. Related Feature - April's Marching Motherboards

Source - OCWorkBench