IMO having a bad stick of ram is going to be one of the last things to be the culprit. RAM manufs. do a series of tests in factory before releasing anything, that's how they come up with the EPP profiles which the mobos can read for "factory overclocking". In my first system, it was just getting the RAM to play nice with the mobo. It was an Nforce4 (I think?) mobo, which historically didn't get along well with some specific RAM sticks by a few of the major manufacturers, because of the components the RAM manuf. used for those specific SKUs. With some finagling it worked.

My second and current system turned out to be more insidious. Long story short, eVGA didn't do a good job of putting a cooling solution on the MOSFETs near the cpu socket. Several were flat out not cooled, the ones that were cooled had bent heatsinks with horrible thermal pads underneath which didn't achieve full contact. I had to manually pull both of these off, straighten them (which was gut-wrenching stressful), and then reattach with a good brand thermal paste. RAM problems went away after that. eVGA caught some real flak from this problem too.

So, LEN, if you still have problems after a week, you'll want to start investigating your mobo and determine if you have cooling issues there. It will deal with either the cooling the northbridge, MOSFETs, airflow over the RAM sticks, or a combination of all of the above.