Scumbria,
>>>Obviously the problem is with the Gateway system, not the K6.<<<
Man, you sure lit up the thread with this one. My 2¢: In I/O interfaces between components or systems from different vendors, sometimes one can be marginally in violation of a timing spec., say the CPU chip vendor in this case, but the other vendor, say the chipset vendor, has enough slop, or grace on their end that the minor violation is overlooked. Everything works. Along comes another chipset vendor which doesn't have the slop or grace, but is closer to the edge of the spec., but still within spec., and bingo, the marginally out of spec. CPU device causes a failure.
Of course you know I'd turn this into an Intel advantage...so many more chipset, DRAM, etc. vendors build around Intel than around AMD, especially early on, that such problems will be found early and fixed in the Intel case.
Further, the much more serious server vendors will be very wary of AMD because of their much lower level of experience (none) in participating in mission critical systems designs. Whether this Gateway thing is AMD's "fault" or not, it can only throw more FUD about AMD into the equation.
Tony |
