I survived 10 hours of Karhu without errors, which I thought was a really good sign. However, every time I restart the PC, I get the motherboard error code A6, and sometimes even on a cold boot. I often have to turn the PC off and on again to get past it.
If I load the XMP profile with default timings, the error disappears completely. Enabling GDM also seems to eliminate the issue (still testing, but so far it looks good).
I changed the ProcODT Pull-Up impedance to 25.3 (the minimum allowed), which improved things for a while, but the A6 code eventually came back.
Any ideas?
2 Comments
Do you pass with Furmark running at the same time? Higher MT/s are very sensitive to SnR on the motherboard’s own traces. I had a “stable” dual rank 7200MT/s set up that would immediately face plant when the GPU was dumping heat and increasing signalling overhead for the motherboard and CPU.
Also your tRFC is *very* tight for M die. Loosen up to 165ns. Likely not the culprit, but my uneducated opinion is that at higher MT/s, you need to loosen up the tRFC time slightly more, even after accounting for the increased MT/s.
It honestly sounds like the motherboard is failing memory training. I would leave termination impedances on auto.
What are your nitro settings? Try to set them looser if you’re occasionally failing memory training.
Well for one you shouldn’t be using 2167 fclk. Secondly you can turn gdm on and see if that resolves it. Gdm really doesn’t make much of a difference latency wise