|[News] AMIGAONE report photographs||ANN.lu|
|List of all comments to this article|
|AMIGAONE report photographs : Comment 42 of 54||ANN.lu|
|Posted by Joe "Floid" Kanowitz on 11-Apr-2002 16:34 GMT|
|In reply to Comment 41 (Amifan):|
The "686B bug" is a little more complex. It's a strange, intermittent interaction between Via's PCI implementation and the controller.
Have a look at this block diagram of an older (KT133A) Athlon chipset using the 686B, noting where the PCI bus "hangs:" http://www.via.com.tw/en/apollo/KT133A.jsp
It's my understanding that the 686B is little more than one large, integrated PCI device. The real problem has been in the north bridges, whose rough equivalent of the Buster chip in an A4000 can't make the PCI bus work properly. Funnily enough, some A4000s had a fairly similar problem: http://www.jeacle.ie/publications/articles/4008/
When not coupled with a buggy northbridge, the 686B should work fine- look at all the boards using AMD chipsets, but 686Bs for IDE, with no reported complaints.
|List of all comments to this article (continued)||