http://www.cs.toronto.edu/~bianca/paper ... rics09.pdf
What's wild to me is all the folks posting that they're going to buy ECC DDR now versus all the overclockers who run their memory out of spec. So many conflicting signals from consumers alone!
DIMM errors in the wild
-
- Grand Pooh-Bah
- Posts: 6722
- Joined: Tue Sep 19, 2006 8:45 pm
- Location: Portland, OR
- Contact:
DIMM errors in the wild
Disclaimer: The postings on this site are my own and don't necessarily represent Intel's positions, strategies, or opinions.
Re: DIMM errors in the wild
I'm accidentally in that crowd. I've had a 1GB dual channel kit rated for 533 lying around for a while, having replaced it with a 2GB of whatever the next step up is (667?). Then I wondered if having 3GB of memory of slower memory would be better than 2GB of faster, so I put the old memory in with the new. Unexpectedly, the BIOS automatically set the clock to the faster rate, so now a third of my memory is operating beyond spec. I figured it wouldn't work, but running memtest and memtest+ overnight for a couple nights in a row didn't show any errors, so I've been willing to trust it. But that amount of testing won't detect errors and uncommon as the paper documents (on the order of one per year). But the paper's conclusions (I skipped the rest) didn't directly address the effects of overclocking, so I'm not sure I've made it worse.
I feel like I just beat a kitten to death... with a bag of puppies.