Personal computing discussed
Moderators: renee, Flying Fox, Thresher
just brew it! wrote:
SNIP
kit even though only one stick was bad, so the system I was building was down for the count.
Flying Fox wrote:But as JBI said there may now be corruptions lurking within. You never know if some corrupted data has been written the registry or files. It is now up to how paranoid you are.
Waco wrote:So you got a newer board that likely is a bit more picky with older higher-voltage DDR3 and you blame the manufacturer? Did you try different speeds, voltages, etc?
Synchromesh wrote:Of course I blame the manufacturer. My memory wasn't overclocked at all, it ran on all-stock settings. That should work regardless of picky boards or anything else. And it did for a year until it went bad.
just brew it! wrote:How do you know for certain that the motherboard isn't going bad instead?
Synchromesh wrote:just brew it! wrote:How do you know for certain that the motherboard isn't going bad instead?
I ran the new RAM they sent me in same slots. It ran memtest overnight with no issues whatsoever. So it's unlikely to be a board problem.
tanker27 wrote:Again, I have to ask if the ram in question is listed on the Mobo manufacturers QVL? If not, then yes, expect some issues.
Savyg wrote:I haven't done any research mind you, but I've never liked the Corsair RAM in the box I was given a few years ago.
I'd replace it but the gains to be seen from faster RAM on the triple channel i7s is pathetic, so why spend the moneys.
I want to though. Eventually.
Synchromesh wrote:It's all DDR3 standard and it should work regardless.
Chrispy_ wrote:Corsair do more testing than the bargain-basement generic RAM so it's still a better bet, but what you're really paying for with Corsair modules is Corsair's lifetime support and RMA system, which is much better than the equivalent service you'd get by buying generic RAM.
Bauxite wrote:Also ECC modules go bad just as much in my experience, it is really for uptime against inevitable bit flips screwing your data or crashing the system, not hardware redundancy. (there is an advanced type of protection for the latter, but only in $$$$$$$$$$$$ kind of systems)
Bauxite wrote:They will only correct for single bit errors, when chips or modules truly fail, they fail HARD. Boot up and the board doesn't see it anymore, or massive errors on anything inside a particular address range.
Bauxite wrote:(there is an advanced type of protection for the latter, but only in $$$$$$$$$$$$ kind of systems)