I have a four-year-old Intel system that has been very reliable, but today
it started having kernel panics. running memtest86 consistently caused a system
reset 47% of the way through test 2.
The system is configured with two sticks of DDR2-800 RAM, 2GB apiece.
I figured I would try each stick alone to isolate the faulty one. Well, memtest86
fails on either stick when used alone, about 12% through test 2.
The motherboard is an ASUS P5Q-EM with four slots. No matter which slot
is used, memtest86 fails. But if I use slot 0, I also get a large, ever-increasing
count of errors in that slot. In the other slot, I just get errors (and the test
gets stuck) but I don't get the per-slot error count.
I had figured that one of the sticks had gone bad and I would replace it.
But they both test bad in exactly the same way. This makes me think
maybe the problem is with the motherboard? But I don't know how to confirm it.
Is there some critical information I should be providing?
If you get the same errors on both DIMMs in all slots then the DIMMs are unlikely to be the problem. As you said, your issue is most likely with the motherboard/chipset but there is also a possibly it could be with the CPU. It could also be a combination.
The only way to know for sure is to proceed by elimination by either testing your components in an otherwise known-good compatible PC or put known-good parts in your PC.
Before you start swapping components, you could try re-seating your CPU. If you are lucky, maybe your problems are caused by a bad contact in the socket that will clear with a simple re-seating.