[conspire] (forw) Re: June 25 MP installfest upgrade RH 7.3 PC to Centos56 continuation

Rick Moen rick at linuxmafia.com
Tue Jul 5 11:42:11 PDT 2011


----- Forwarded message from Rick Moen <rick at linuxmafia.com> -----

Date: Tue, 5 Jul 2011 11:41:47 -0700
From: Rick Moen <rick at linuxmafia.com>
To: wood eddie <ewood111 at yahoo.com>
Subject: Re: June 25 MP installfest upgrade RH 7.3 PC to Centos56
	continuation
Organization: If you lived here, you'd be $HOME already.

Eddie, reviewing a previous mail made me notice something else, probably
nothing worth worrying about (a bit too late anyway), but I thought I 
should mention it:

> Thanks for reply. Now I am changing my PC hardware. Today 
> bought Adaptec 2940 SCSI card, memory upgrade 512MB to 2 GB
> [1 GB PC2700 + 1 GB PC3200 memory].
        ^^^^^^        ^^^^^^

Hmm.  Mismatched memory is somewhat undesirable, and should be avoided
if reasonably possible.  As with many other things in computing, it's an
area where you can get away with playing fast and loose _most_ of the
time (add memory that's faster but of the same class), but your system
_can_ be plagued by subtle memory-timing problems (rare random freeze-ups
or reboots).  It shouldn't happen, but remains a minor possibility.

You might think that 'Oh, my system has 1 GB of 168 pin double data rate
DDR333 aka PC2700 non-ECC SDRAM.  Since I'm having a difficult time
finding more PC2700, and the world has moved on to faster RAM, I'll 
just add a gig of DDR-400 aka PC3200 non-ECC SDRAM.  Since it's better
and nimbler RAM, that should be fine, since it can keep up with the CPU
and memory bus _even better_ than the existing RAM can.'

Usually that works -- and faster DDR SIMMs are _intended_ to be
compatible with (backwards compatible to) slower DDR SIMMs.[1]  However,
if you want to be cautious, you should match RAM as closely as possible
whenever expanding system memory.

[1] See, for example, http://www.crucial.com/support/memory_speeds.aspx


----- End forwarded message -----




More information about the conspire mailing list