[gmx-developers] Coulomb LR energy totals on AMD???

[David van der Spoel]

On Tue, 2003-09-09 at 16:06, Anton Feenstra wrote:
> Hi,
> 
> 
> We've seen a reproducible error here on our 32 single CPU AMD 1G cluster,
> which appears *only* in the Coloumb LR energy totals (both in .log and .edr).
> The values are bogus, jumping from e.g. -1e04 to -1e11 or even -1e35!
> Averages are -inf, rmsd nan. Simulations run for several ns, so it appears
> forces are not affected. These values do not occur on SGI, or our dual
> 1.3G AMD machines. It therefore smells like a nasty bug!
> 
If it is amd only, then I presume that it has to be in one of either
PME, fftw or the ewald corrections. If you say it is reproducible do you
mean that you get the same weird result with the same run repeatedly?

The line of testing should be first to turn off assembly in compiling
(this will also turn off some other assembly bits in pme.c), then, if
the error persists use CFLAGS=-g, a further option could be to upgrade
your fftw (2.1.5) or to compile fftw without the i386 hacks..


-- 
Groeten, David.
________________________________________________________________________
Dr. David van der Spoel, 	Dept. of Cell & Mol. Biology
Husargatan 3, Box 596,  	75124 Uppsala, Sweden
phone:	46 18 471 4205		fax: 46 18 511 755
spoel at xray.bmc.uu.se	spoel at gromacs.org   http://xray.bmc.uu.se/~spoel
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++