[Beowulf] Nehalem memory configs
cap at nsc.liu.se
Thu Apr 9 00:35:52 PDT 2009
On Wednesday 08 April 2009, Jan Heichler wrote:
> JL> Peter Kjellstrom wrote:
> >> Can anyone confirm that mixing different sizes of dimms, even when
> >> keeping to a by-three-symmetric configuration, actually does degrade
> >> performance?
> >> That is, first hand information that config 1 will be slower than config
> >> 2:
> >> 1: two sockets, 3x 1G and 3x 2G per socket RDIMM (total 18G @1066)
> >> 2: two sockets, 6x 1G(or 2G) per socket RDIMM (total 12G (or 24G) @1066)
> >> /Peter
> JL> I can confirm it because I've got it on a test box I've been running :)
> Can you give us the result of a stream-benchmark? with 12 x RDIMM in
> asymmetric config and 6 x RDIMM ? both running at 1066?
Note that both my configs 1 and 2 above are 12xRDIMM and "symmetric". The only
thing that differs is wether it's purely one DIMM size or two.
12x vs 6x @1066 is a completely different config and I wouldn't be surprised
to see a performance difference.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Size: 189 bytes
Desc: This is a digitally signed message part.
Url : http://www.scyld.com/pipermail/beowulf/attachments/20090409/7d5e866b/attachment.bin
More information about the Beowulf