Page 1 of 1

VASP 6.4.1 test failures: HEG_333_LW SiC8_GW0R SiC_ACFDTR_T

Posted: Tue Jun 18, 2024 10:20 pm
by michael_walkup
I successfully compiled VASP using an Intel 2018a module in Ubuntu, and the test returns failures for HEG_333_LW, SiC8_GW0R, and SiC_ACFDTR_T.

For all three, the energy values the log states the calculated do not match the reference values, and the energy file is empty. SiC8_GW0R additionally has the GW output yield different results, and the eigenvalue files do not have the same number of rows as the reference. I can supply any of these files if needed. Attached is a zip file of makefile.include and testsuite.log. Are these errors significant, and is there some parameter not correctly set up in the makefile?

Thanks,
Michael

Re: VASP 6.4.1 test failures: HEG_333_LW SiC8_GW0R SiC_ACFDTR_T

Posted: Thu Jun 27, 2024 12:39 pm
by pedro_melo
Dear michael_walkup,

I ran the testsuite for VASP 6.4.1 on one of our local machines and all tests passed successfully. Could you send me the files that are reported as different in testsuite.log? For example, in your SiC8_GW0R test it is reported that

Code: Select all

 Comparing files: g0w0_eigenvalues and g0w0_eigenvalues.ref
                           288  number(s) differ.
Same for the other tests. This could be something related to your installation or local definitions.

In principle there is nothing that should prevent you from running VASP 6.4.1 for your normal calculations, as this does not seem to indicate that there is some bug or unidentified error.

Kind regards,
Pedro

Re: VASP 6.4.1 test failures: HEG_333_LW SiC8_GW0R SiC_ACFDTR_T

Posted: Thu Jun 27, 2024 7:57 pm
by michael_walkup
Hello Pedro,

Attached is a folder containing the incompatible files, as requested. Thank you very much for looking into this.

-Michael