is it mobo or CPU memory controller problem

[tb2496]

I've been having memory problems for a while. after doing the usual troubleshooting I thought the problem is most likely the motherboard, based on these findings:

1. memory errors happen after the PC is running for a while, could be days

2. memtest gets errors always around the same address ~5600GB when dual channel and interlaced is enabled, and ~7100GB when dual channel and interlaced are disabled

3. underclocking CPU, changing memory timing or voltage doesnt help

4. even if I move DIMMs around the errors happen at the same place which would indicate its not a faulty DIMM

5. if I power down and up PC the problems go away and start again after some time

AMD Athlon II X2 250
4 x G.SKILL F3-12800CL9D-4GBNQ PC3-12800
Gigabyte GA-880GM-UD2H AMD880G

one thing I'd like to know is when I enable/disable dual channel and interlaced in bios is that changing something on the mobo or the CPU's memory controller?

[Tradesman]

I'd guess the MC is getting stressed, may need additional voltage to CPU/NB (MC (memory Controller) or DRAM itself. Would prob also help to drop to 1333, your CPU isn't really good got 1500 sticks

[tb2496]

I'd guess the MC is getting stressed, may need additional voltage to CPU/NB (MC (memory Controller) or DRAM itself. Would prob also help to drop to 1333, your CPU isn't really good got 1500 sticks

I've tried both things, I run the memory at 1333 (tried going below as well) but problem came back.

[Tradesman]

Can try each stick individually to check they are good, then would check CPU and socket for debris, excess thermal compound on bottom of cpu or in socket and check closely for bent/broken pins

[tb2496]

Can try each stick individually to check they are good, then would check CPU and socket for debris, excess thermal compound on bottom of cpu or in socket and check closely for bent/broken pins

thanks for the suggestions but I'm not sure that applies to my situation. I would imagine the problems would be more constant and not have the specific patter I see.

I'll continue troubleshooting though.

[Tradesman]

A bent or broken pin can directly affect the MC (memory controller), am just wondering if that is why you get the same errors at the same points, especially if moving sticks around, if they were in different slots at different testing they should pop errors at a different memory address if a stick is bad

[tb2496]

A bent or broken pin can directly affect the MC (memory controller), am just wondering if that is why you get the same errors at the same points, especially if moving sticks around, if they were in different slots at different testing they should pop errors at a different memory address if a stick is bad

if it was something with a pin then wouldnt the problem be there all the time?

I think I might have checked for that, I might have also checked the DIMM slots couple months ago (when I first noticed something wasnt right and started troubleshooting). after trying different things I concluded that it was my 3rd dimm slot because problems disappeared when I left it empty. things seemed stable until now, when even with that slot being empty errors started happening.

last couple weeks I've tried again running with 4 dimms and errors showed up in the same address (but sometimes for 5 days things are fine, other time after 7 hours of uptime errors happen). now I went down to 2 dimms and will see what happens.

[Tradesman]

We'll be waiting to hear