In message <20060308113557.5a8a66bb.segv@netctl.net>, segv@netctl.net writes:
>On Tue, 07 Mar 2006 20:22:42 -0600
>seebs@plethora.net (Peter Seebach) wrote:
>
>> In message <20060308020626.GA307@gallia.cubidou.net>, Quentin Garnier writes:
>> >The memory is probably failing.
>> 
>> I don't think so.  memtest86 can run on the hardware for days without a single
>> error, and, once again, it is happening on EVERY system I have access to to
>> test on.  It would be very odd to me for three machines to all have "failing"
>> memory that fails under a specific load...

>Is there something in common between those machines? The same manufacturer and
>chipset, or something like that?

All of them are nForce4 chipsets.  I'd love to see tests on something else,
but I haven't got anything else compatible to try on.  On the other hand, so
far as I know, the memory controller is on-chip on these systems, so it
shouldn't matter.  Probably.

-s