So what exactly causes mem controllers to die?

[Matt2]

I've been hearing off and on about AMD memory controllers dying and thus making a nice $400 paper weight.

What causes this? Who's at risk? How often does this happen?

 

[Jeff7181]

Since it's built into the CPU core now, I'd assume anything that could normally kill a processor could also kill the memory controller.

 

[alpha88]

I have never heard of a CPU dying, or it's memory controller (intel or AMD), other than:

No heat sink
Wild overclocking
Crushed core

 

[Unkno]

Heat and/or voltage will kill it...

 

[Stoneburner]

alzheimers?

 

[PingSpike]

There was a good and lengthy thread about this on ocforums...the consensus was high amounts of voltage to ram was believed to be the cause. I believe greater then 3.2?

 

[sumyungai]

Originally posted by: PingSpike
There was a good and lengthy thread about this on ocforums...the consensus was high amounts of voltage to ram was believed to be the cause. I believe greater then 3.2?

I have heard the same thing but still don't understand how high voltage going to the RAM is going to wear down the memory controller that is located on the CPU that is not receiving the high voltages.

 

[PingSpike]

Originally posted by: sumyungai

Originally posted by: PingSpike
There was a good and lengthy thread about this on ocforums...the consensus was high amounts of voltage to ram was believed to be the cause. I believe greater then 3.2?

I have heard the same thing but still don't understand how high voltage going to the RAM is going to wear down the memory controller that is located on the CPU that is not receiving the high voltages.

I believe this is the thread I was thinking of. It has a better explaination then I can offer. But the basic idea is that since the memory controller is on the cpu die, the memory voltage is run through there as well. The theory is it has to do with a disparity between the memory voltage and the cpu voltage causes more stress then the high voltage itself.
Text

 

[Zap]

Originally posted by: PingSpike
The theory is it has to do with a disparity between the memory voltage and the cpu voltage causes more stress then the high voltage itself.

That's what I've heard as well. The thinking is that with super high memory voltages, if you undervolt your CPU (manually or through Cool and Quiet) then the difference is what kills. Supposedly people were dared to try it and some did, and their CPUs died after a few reboots.

 

[jazzboy]

I'm a bit concerned myself, because on one of my computers I'm running the cpu undervolted (Newcastle 3000+ undervolted to 1.35) with ram at 2.65v.

And then what Cool n Quiet as that set all A64s to 1.1v when idle/light use?