[QE-users] [SPAM] QE7bug

[Paolo Giannozzi]

Actually the "bug" (sort of) was in the 6.7 version, not in the most 
recent ones. The code fails to converge (you set a very small threshold 
for pressure) and stops after resetting the history twice (history is 
reset when the optimization procedure runs into trouble). In v.6.7, it 
stops claiming convergence (not true); in later versions, the code stops 
with no convergence (technically true)

Paolo

On 04/09/2023 02:39, 526587466--- via users wrote:
> Dear Developers,
> 
> Recently, I did a test about the geometry relaxtion of graphene by 
> different QE version (i.e., 6.7, 7.1 and 7.2 version).  Only the 6.7 
> version gives the convergened result, while not for the 7.1 and 7.2.  
> Please see the attachment.
> 
> Is there a bug or other things in the QE 7 version?
> 
> Thanks
> 
> Roc
> 
> Henan Normal University, China

-- 
Paolo Giannozzi, Dip. Scienze Matematiche Informatiche e Fisiche,
Univ. Udine, via delle Scienze 206, 33100 Udine Italy, +39-0432-558216