Dear Dr.Giuseppe and Dr. Lorenzo,

The calculation has successfully completed on my local server without any
error.
However, I am not yet sure why I am getting the error on Supercomputer, and
how to get rid of it.

Thank you.


On Mon, Oct 23, 2017 at 9:54 PM, B S Bhushan <[email protected]> wrote:

> Dear Dr. Giuseppe,
>
> I have already submitted the same job on a different machine (my local
> server) and awaiting the result. So far I have not seen any error, yet I am
> awaiting the final result.
>
> The error I posted was actually received when I tried to perform it using
> a High Performance computing facility.
>
> I will let you know of the result from my local server.
>
>
> On Mon, Oct 23, 2017 at 9:48 PM, B S Bhushan <[email protected]> wrote:
>
>> Thank you very much for your support Dr. Lorenzo.
>>
>> I will try these and let you know the result.
>>
>> On Mon, Oct 23, 2017 at 9:07 PM, Lorenzo/Eric Paulatto <
>> [email protected]> wrote:
>>
>>> Dear B S,
>>>
>>> I've had a look at your input, I see that the C-N distance is already
>>> quite small at the beginning, and become even smaller after a couple of
>>> iteration; definitely smaller that the core radius of the two
>>> pseudopotentials used, which is a typical cause of diagonalisation problems.
>>>
>>> You should try using some harder pseudopotentials, I've attached two
>>> hard pseudopotential fro mthe pslibrary of Andrea Dal Corso (do not forget
>>> to cite), which should do the trick. Take care of converging the cutoff.
>>> UPDATE: I'm not attaching the pseudos because they don't pass in the list,
>>> I'm sending them privately in a minute
>>>
>>> Please let us know if this solves your problem!
>>>
>>> kind regards
>>>
>>> On 23/10/17 15:08, B S Bhushan wrote:
>>>
>>> Dear Dr. Lorenzo,
>>>
>>> I think the error is occurring in the 4th optimization step.
>>> I have attached the output file for your reference.
>>>
>>>
>>> Awaiting your kind response.
>>> B S Bhushan
>>> Ph.D Scholar,
>>> ABV-IIITM Gwalior, India.
>>>
>>>
>>>
>>> On Mon, Oct 23, 2017 at 4:44 PM, Lorenzo Paulatto <[email protected]>
>>> wrote:
>>>
>>>> is it at the first structural relaxation step or later? I'm afraid that
>>>> doing a vc-relax of a small supercell with a defect could cause an
>>>> unphysical crunch
>>>>
>>>> On 21/10/17 18:52, B S Bhushan wrote:
>>>>
>>>> Dear Expert,
>>>>
>>>> I was getting the below error while trying to vc-relax a doped graphene.
>>>> can you please suggest why is it occurring??
>>>>
>>>>
>>>>      iteration #  9     ecut=    30.00 Ry     beta=0.70
>>>>      Davidson diagonalization with overlap
>>>>
>>>>  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>>>> %%%%%%%%%%%%%%%%%%%
>>>>      Error in routine  cdiaghg (261):
>>>>       problems computing cholesky
>>>>  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>>>> %%%%%%%%%%%%%%%%%%%
>>>>
>>>>      stopping ...
>>>> application called MPI_Abort(MPI_COMM_WORLD, 1) - process 22
>>>> application called MPI_Abort(MPI_COMM_WORLD, 1) - process 20
>>>> application called MPI_Abort(MPI_COMM_WORLD, 1) - process 4
>>>> application called MPI_Abort(MPI_COMM_WORLD, 1) - process 2
>>>>
>>>>
>>>>
>>>>
>>>> Sincerely,
>>>>
>>>> B S Bhushan,
>>>> Ph.D scholar,
>>>> ABV-IIITM Gwalior, India.
>>>>
>>>>
>>>> _______________________________________________
>>>> Pw_forum mailing 
>>>> [email protected]http://pwscf.org/mailman/listinfo/pw_forum
>>>>
>>>>
>>>> --
>>>> Lorenzo Paulatto - Paris
>>>>
>>>>
>>>> _______________________________________________
>>>> Pw_forum mailing list
>>>> [email protected]
>>>> http://pwscf.org/mailman/listinfo/pw_forum
>>>>
>>>
>>>
>>>
>>> _______________________________________________
>>> Pw_forum mailing 
>>> [email protected]http://pwscf.org/mailman/listinfo/pw_forum
>>>
>>>
>>> _______________________________________________
>>> Pw_forum mailing list
>>> [email protected]
>>> http://pwscf.org/mailman/listinfo/pw_forum
>>>
>>
>>
>
_______________________________________________
Pw_forum mailing list
[email protected]
http://pwscf.org/mailman/listinfo/pw_forum

Reply via email to