[Wien] IBM Blue Gene

Peter Blaha pblaha at theochem.tuwien.ac.at
Sat Sep 24 11:01:35 CEST 2011


As far as I know from tests on a Blue Gene (about 4 years ago), WIEN2k 
will not run "out of the box", since the Blue gene does not support 
scripts, and has very little memory/node, so that even trivial programs 
like mixer may not fit for big cases.

We tested lapw1 on up to 1000 nodes. Scaling was ok, but the overall 
speed of those processors was very slow, so that 100 Intel Xeon cores 
ware much faster than 1000 BG.

Unless you have a big amount of free cpu-hours and are willing to spent 
some time for implementation (and have the knowledge), I would not go to 
such a "specialized" machine.

Am 23.09.2011 18:23, schrieb Lee, Yongbin [A LAB]:
> Dear All,
>   Is there anyone who has run WIEN2k on an IBM Blue Gene
> and can tell how the performance is(was) ?
>
>    Yongbin
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien

-- 
Peter Blaha
Inst.Materials Chemistry
TU Vienna
Getreidemarkt 9
A-1060 Vienna
Austria
+43-1-5880115671


More information about the Wien mailing list