Timo Schoeler wrote:
> hi,
> 
> i think that steve is on the right path regarding to this topic; to me 
> it sounds very likely to be a hardware related issue, i.e. steve said 
> that the combination of both dimms caused trouble seems to be a problem 
> with the second dimm's added load (capacity, timing etc.) to the 
> (memory) bus.
> 
> memtest86 could provide a hint.
> 
> is anyone able to test a machine with the same config (i.e. same memory 
> size), but using *ECC* ram? maybe that would kick the bucket... ;)


   Sorry if that got lost in the thread, but I *am* using ECC DIMMs.
They're the standard Compaq parts for this system.  I guess I'm
only assuming it's *doing* ECC with them.  I'm assuming that's
actually the issue...  The ECC/partity is causing the NMI.  Unless
it's something in the software that's misusing the memory somehow,
but.

   Again, I'll see what a current tree runs like.  Then if I'm
still having trouble, I'll likely try memtest86 and/or memtest86+.

                                - Chris