Via has finally fessed up, according to The Register. After a bit more investigation into the impact this problem has on different north bridge chips to which the 686B can be attached, they will release a corrective BIOS to manufacturers.
"VIA's BIOS fix works by adjusting a number of PCI settings, which, according to TechChannel, suggests the problem is a result of competitive PCI access."
Many of you who will read this will likely experience deja vu. Threads on various forums on the Internet will quickly confirm this problem is not in the mind of a few. The HardForums at HardOCP have a very large and growing thread here. It started with a fair amount of skepticism and has grown from there as others became acutely aware the sitaution was not caused by incompetence. ViaHardware's forum currently has a beast of a thread detailing issues with the SBLive! and the Via chipset.
As the HardForum thread and the Register both mention, this problem is exacerbatedbut not causedby SBLive! sound cards. Many find relief from the problem by removing their SBLive! cards from their PCs. Others find the problem never quite goes away.
I've indicated my displeasure with the Via south bridge chip before. Due to address line problems with my own 686B-based motherboard, I became skeptical of Via's ability to properly implement the PCI spec. I guess that I wasn't too far off the mark. I wonder what kind of performance hit this fix will cause.
It amazes me that people still implement Via chipsets in mission-critical environments. Via is still, at best, an enthusiast chipset manufacturer. (Thanks to TwoFer for the link.)