Can SCTK deal with 2D materials?

[mitsuaki1987]

(2021-03-23 11:34 by DaMing_1566 )
Dear developers,

Recently, I calculated the superconducting properities of doped graphene using lambda.x in Quantum Espresso and wanted to examine that using SCTK.
After calculating the screened Coulomb interaction, the kel.out file read

Intel MKL ERROR: Parameter 10 was incorrect on entry to ZGEMM

and I could not figure out why it happened ( In all of the calculations, I used $k_z$=1, is it neccessary to increase the number of $k_z$ ).
Can SCTK deal with 2D materials?

It will be greatly appreciated if you could spare time to answer my questions.

Yours, Lu Qing

[mitsuaki1987]

(2021-03-26 01:10 by mitsuaki1987)
Dear Dr. Lu Qing

Thank you for reporting. I will try the graphene case.
Probably $k_z$ = 1 is OK.
But Could you also try $k_z$=2?

Best regards,
Mitsuaki Kawamura

[mitsuaki1987]

(2021-08-31 10:31 by pfliu #87974))
Hi, Mitsuaki Kawamura

I had the same problem with the errors of

Intel MKL ERROR: Parameter 10 was incorrect on entry to ZGEMM

for 2D metals with $k_z$=1.
I have also tried $k_z$=2, however, it does not work.
I checked the $kel.out$ of 2D metals and found unusual output as follows:

##### FFT wfc(G) -> wfc(r) #####

Grid : 1 1 1

I modified part of the code sctk_wfc.f90

DO ii = 1, 3
  igmin = MINVAL(igv(1:3, 1:npwxtot, 1:cnt, 1:2)) 
  igmax = MAXVAL(igv(1:3, 1:npwxtot, 1:cnt, 1:2))
  WRITE(stdout,'(7x,"igmax :",2(2x,i0))') igmax,igmin
  CALL mp_max(igmin, world_comm) 
  CALL mp_max(igmax, world_comm) 
  WRITE(stdout,'(7x,"igmax :",2(2x,i0))') 
  igmax,igmin nf(ii) = good_fft_order(igmax - igmin + 1)
END DO

For 2D metals, it outputs the following content:

##### FFT wfc(G) -> wfc(r) #####
igmax : 43 -42 
igmax : 43 2147483647 
igmax : 43 -42 
igmax : 43 2147483647 
igmax : 43 -42 
igmax : 43 2147483647 
Grid : 1 1 1

For 3D Al, it outputs the following content:

##### FFT wfc(G) -> wfc(r) #####
igmax : 6 -7 
igmax : 6 -7 
igmax : 6 -7 
Grid : 15 15 15

I have no idea what I should do.
I wonder if you have updates of the code.
Thank you and I am looking forward to hearing your reply.

Sincerely, Liu.