faulty memory

[Stijn Hoop]

On Mon, Sep 22, 2003 at 02:50:16PM -0400, Lowell Gilbert wrote:
> Stijn Hoop <stijn at win.tue.nl> writes:
> > I just had this problem, P4 on Gigabyte mainboard, DDR400 memory, 2 DIMMs
> > of 512 MB each. With a single DIMM (both DIMMs tested) memtest didn't
> > find anything and the machine was rock stable, but in any combo of the
> > 2 DIMMs, dual channel or not, the machine would lock up / throw
> > gcc ICE everytime.
> > 
> > When I ran memtest on the machine with both DIMMs in, it would consistently
> > show errors in one 512 half though, and so I returned the one 'faulty' DIMM
> > and all was well.
> > 
> > Talk about strange...
> 
> Not that strange.  The second socket was having problems.  
> Maybe just couldn't handle the capacitance load it was 
> getting (which could be different with other modules; 
> other sized modules, anyway).

Err, my sentence above tried to convey that, with both DIMMs in the machine,
running memtest produced consistent errors _IN ONE DIMM_. That is, I swapped
the two DIMMs and memtest started showing errors in the other 512MB. 

In fact, the machine is running happily with the swapped DIMM in the same
socket that the faulty one had. So I doubt that it's the socket :)

--Stijn

-- 
"Linux has many different distributions, meaning that you can probably find
one that is exactly what you want (I even found one that looked like a Unix
system)."
		-- Mike Meyer, from a posting at questions at freebsd.org
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 187 bytes
Desc: not available
Url : http://lists.freebsd.org/pipermail/freebsd-chat/attachments/20030922/7a173330/attachment.bin