So it's been a day since the Rowhammer issues reared its ugly head.
>What the fuck is a row hammer?
https://en.wikipedia.org/wiki/Row_hammer
How are your DIMMs fare /g/?
Did they pass or were they susceptible?
Latest memtest has a test for it (test #13)
http://www.memtest86.com/download.htm#beta
>early adopter problems: the thread
cucks
>>53551140
So you're still using DDR2?
>>53551147
SRAM, bitch!
Does this effect ECC setups?
>>53551127
>So it's been a day since the Rowhammer issues reared its ugly head.
Didn't they disclose this problem months ago?
>>53552154
>>53552327
Actually, yes it can corrupt even with ECC. (as far as I have read)
This is why you supersample with foil...
>>53551147
>newest RAM in my desktop and laptops is DDR2
>I'm still using VGA for my monitor
>my newest computer (my desktop) turned 7 earlier this year, though some parts of it are newer
>using a 1680x1050 LCD from 9 or 10 years ago
>finally stopped using IDE storage devices in December
>my floppy drive died last year
>>53552154
It can, ECC memory can only correct 1 bit per row, if more than one bit in a row is flipped it cant be reverted by standard ECC ram.
>>53551127
I see this old post here which offers a way to get around the problem:
http://www.overclock.net/t/1540772/cmu-intel-ed-et-al-large-portion-of-modern-dram-vulnerable-to-row-hammer-errors/10#post_23529304
Good thing I use ddr3 ecc
>>53553138
>>53557207
I thought parity could only fix one bit per row.
>>53551127
>two PC notebooks are C2D / DDR2
>oh shit what about muh MacBook Pro!
>look it up
>Apple fixed the problem a while back with firmware update that increases row refresh
Feels good.
>>53557207
It may not be able to correct more than a bit, but it can still invalidate the entire row.
