[QE-users] Error while NSCF calculation

[Akhil g.nair]

Dear all I've encountered an error while conducting an NSCF calculation on an SCF-run structure. The error message I'm encountering is as follows:

"
     Error in routine cdiaghg (2776):     S matrix not positive definite"
 This error consistently appears after any kpt#. I've attempted various troubleshooting steps, including adjusting the number of processors involved in the calculation. Surprisingly, increasing the processor count leads to early NSCF failure. Similarly, attempts to mitigate the issue by increasing the values of Ewfc and Erho result in NSCF failure at an earlier kpt#.
I've experimented with reducing the processor count, but this only delays the failure to a higher kpt#. Additionally, decreasing Ewfc and Erho values prevents SCF calculation convergence.
I would greatly appreciate any insights or potential solutions you may have regarding this matter.
RegardsAkhil NairResearch scholarHBNI, INDIA
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20240307/dbf23bda/attachment.html>