----- Original Message ----- 
From: "Jeff Waugh" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Friday, 18 January 2002 10:10
Subject: Re: [SLUG] Putting faster RAM in a machine that can't deal with it


> <quote who="David Kempe">
> 
> > This has worked for me on a number of boards - even putting 256Mb RAM in a
> > old VX board with a P200 I think worked but only counted 128Mb.
> 
> Almost; those motherboards would only support caching up to 64mb of RAM.
> More than that, and they'd turn it off completely. For some applications, it
> meant that more RAM actually killed performance (I was doing 3d modelling
> back then, and it was certainly true in that case). You had to have an HX
> chipset if you wanted decent performance with lots of RAM.
> 

Nope, I have seen pretty much the same thing happen on many older boards
sporting DIMM sockets - sometimes less ram is detected than the actual
DIMM size, sometimes it refuses to work at all.
(Not that I am saying the caching issue is not real, but that there are
real compatability problems with older implementations of DIMM controllers)..


For instance, I had bought several rated 128MB DIMMs a couple of years ago.
These refused to be detected at all in several old Digital desktop machines
(Petium 233MMX vintage), though the same machines happily run with 128MB
DIMMS from different manufacturers. The new DIMMS worked fine in my BX
motherboard (as well as serveral other BX motherboards) - All detected
as 128MB.

Jump forward a year or so....

Put the same DIMM into a Athlon system, and suddenly the DIMM is seen as
256MB! (they were/are mislabled 256MB DIMMS?!).


> I *thought* we didn't have stupid bollocks like that going on these days.

I suspect this is only true if all your hardware comes from the same vendor
(Sun, Apple, etc)...

--


-- 
SLUG - Sydney Linux User Group Mailing List - http://slug.org.au/
More Info: http://lists.slug.org.au/listinfo/slug

Reply via email to