[conspire] Parts is Parts

[David Fox]

On Thu, Apr 10, 2008 at 12:15 PM, Rick Moen <rick at linuxmafia.com> wrote:

>  Probably not crucial -- but I've often wondered how much bad RAM is in
>  use that its users have no idea is bad.  I also have to wonder about
>  quality loss with falling prices and rising RAM densities.

Indeed. And many of the "casual" uses don't catch issues, especially
overclocked boxes. People think they've got the killer game
workstation that's fast as anything but don't try to run something
like prime95 (aka mprime, or GIMPS - the Great Internet Mersenne Prime
Search) whiich is good at detecting problems, although it isn't
designed of course per se to do that - it's just the enormous amount
of calculations it's got to do.

The "torture test" option does a credible good job at exercising RAM,
power supply, cpu/cache etc. But I'm not sure how well it exercises
the whole of one's RAM, although it does some longish interval (like 1
meg spans) fft tests.

>  Back around 1995 when the Intel Triton motherboard chipsets came out
>  lacking memory parity support (and were a big success because of the

I thought the same thing, having been use to parity ram on my first xt
and later class boxes. I used to sneer at some of the other people
with RAM that didn't have any parity checking - especially given
(then-big) memory spaces one would come to the conclusion that it was
certainly needed.

>  Note that bad RAM can have a much more pernicious effect than does bad
>  storage media:  The firmware and OS-level routines for testing and

Excellent point.

>  You want to keep increasing the number while watching the "si" and "so"
>  fields on /usr/bin/vmstat's report of swap activity:  Those are the

Aha. I'll keep that in mind for the next time.