[mesa-users] off-center Ne and O burning (again…)

[Francis Timmes]

hi falk,

>  Here is an example: a while back when we were switching from intel 11 to 12 compilers we checked the impact on AGB evolution. The same version of MESA would produce different result in terms of Teff and L on the advanced AGB. In the end I believe it was related to overly aggressive optimization of floating point arithmetic in intel 11.

could be my leaky memory, but i recall ifort 12.x significantly loosened 
the floating point behavior of various operations for speed (ifort 12
is twice as fast as ifort 11!). this was the genesis for specifying a 
small bevy of ifort compiler options aimed at controlling the floating point behavior. 
i recall it helped but didn't completely solve the AGB issues we were seeing.

fxt