I found another issue with ifx that appears in EOSFIT on AMD processors. When using ifort from OneAPI 2024.2, eosfit runs without problems on AMD processors, however, the fitting of all eos curves fails (eos2 bombs NANs) when it was compiled with -O2 and ifx from OneAPI 2024.2 or later (I did not check older Versions of ifx, that were even more buggy). It works with -O0 or -O1.
Indeed, Intel may not be much interested that any code runs well on EPYCs and as good as on XEONs Did anyone use the AMD compiler suite and libraries to compile Wien2k ? or gnu fortran with the AMD libraries ? I checked lmdif.f but this code is very old fashioned, does someone know a more modern version, just to test if it causes an "optimisation" problem Ciao Gerhard DEEP THOUGHT in D. Adams; Hitchhikers Guide to the Galaxy: "I think the problem, to be quite honest with you, is that you have never actually known what the question is." ==================================== Dr. Gerhard H. Fecher Institut of Physics Johannes Gutenberg - University 55099 Mainz _______________________________________________ Wien mailing list [email protected] http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien SEARCH the MAILING-LIST at: http://www.mail-archive.com/[email protected]/index.html
