RAM overheating?

[Dave3000]

About 1 week ago I upgraded to i7 3930k / Asus P9X79 Pro and a Hyper 212 Evo CPU cooler. I noticed that one of my memory sticks went bad, as a few days ago my system started freezing in Alan Wake, BSOD during Windows 7 downloading updates, and App Crash messages using Internet Explorer. All this happened all of a sudden, came from stable to unstable without warning, and started during playing Alan Wake a few days ago. I was able to find out that one my my memory sticks went bad as I tested each stick with Memtest 86+.

Now I'm wondering why one of my memory sticks suddenly goes bad on me. Could it be the hot air from my CPU cooler the blows over the memory toward the exhaust fan inside my case(Socket 2011 motherboards have DIMM slots on both sides of CPU socket)? Could it just have been bad luck? Could it have been my motherboard destroying my memory, although I can't really blame it on the motherboard right now unless it's going to happen again?

Would going for liquid cooling prevent this memory failure from happening? Problem is that I tried an H80 and had a difficult time fitting the bracket holes through the standoffs and it made too much noise to powering on my PC for the first 15 seconds. I tried installing an H60 but it wouldn't fit my socket as one of the holes of the bracket didn't line up with one of the standoffs and had to return that one as well. So I'm back to my Hyper 212 Evo and hoping that the heat from the Cooler won't destroy my RAM.

 

[Dave3000]

I want to add that before upgrading to my i7 3930k, I ran my DDR3 1600 1.65V memory at the auto 1333 @ 1.5v on my i7 2600k system since Sandy Bridge only officially supported up to DDR3 1333 @ 1.5v and I didn't run the memory at it's full 1600 speed for the majority of the time when I was using it in my i7 2600k system since it wasn't within specifications of that CPU.

 

[T_Yamamoto]

maybe your ram is at 1.65 v

 

[bryanl]

Would going for liquid cooling prevent this memory failure from happening?

Using high quality name brand memory, like Samsung, would prevent the failure from happening. I suspect you currently have unbranded memory chips.

 

[Dave3000]

Using high quality name brand memory, like Samsung, would prevent the failure from happening. I suspect you currently have unbranded memory chips.

Actually, they were good quality. They were Kingston HyperX DDR3 1600 1.65v. I'm not sure of the brand of memory chips on the modules though since a heatsink attached to the module. I did get an advanced RMA from the manufacturer. However, I still went ahead and bought 16GB of Corsair Vengeance DDR3 1600 1.5V low profile because I kept reading on the internet that it's recommend to not use higher than 1.5V memory on Sandy Bridge chips and I don't want to risk frying my $600 CPU if I want to run my DDR3 1600 at 1600MHz.

 

[bryanl]

Would going for liquid cooling prevent this memory failure from happening?

Using high quality name brand memory, like Samsung, would prevent the failure from happening. I suspect you currently have unbranded memory chips.

Actually, they were good quality. They were Kingston HyperX DDR3 1600 1.65v. I'm not sure of the brand of memory chips on the modules though since a heatsink attached to the module.

I hate to say it, but they are not good quality, or else they would not rated 1.65V. The brand of the module matters very little, compared to the brand of its memory chips, and with all quality DDR3 modules except Samsung, the chip brand can be identified by merely looking at them. It's neither recommended nor ever necessary to use higher than 1.50V for DDR3, except if the memory chips are defective.

 

[Dave3000]

Well I read on the internet that you shouldn't buy generic memory if you can afford to avoid it. You can get generic memory modules with good memory chips as I have bought a generic memory with Micron memory chips on it many years ago, the store advertised it as Micron memory but only the chips were Micron brand. Memory chips aren't all that make a good quality DIMM, as quality of the module is important as well. I've always had a good experience with Kingston memory, except for a week ago when one module went bad on me, and the RMA process went smoothly with an advanced RMA replacement.

I had a good experience with Kingston ValueRam as well, as I don't overclock and they ran stable at their advertised specs and they conformed to JEDEC specifications. My dad had a 256MB Corsair ValueSelect memory stick, which was the most basic memory conforming to JEDEC specification from Corsair, go bad after a few years.