a single localDeviceDOS simulation took forever to run

[yang su]

Hi

I tried to run a single local device DOS simulation which involves around 200 atom on a super computing machine and I set the walltime to be 48hrs. However, it took the entire 48 hrs and still did not finish. I hope it did converge but I am not sure. Can you take a look at my script and tell me what is the problem? Thank you very much

[Anders Blom]

The simulation cell is very large, and most importantly contains a huge amount of useless vacuum. You should most definitely remove the vacuum above and below the spatial regions, and try with that, but you will still perhaps need to reduce the spatial regions themselves.

[yang su]

Thank you very much for your help and I finally had the job finished and got the nc file. Now I would like have a band diagram analysis of the device. How can I plot the band diagram(like band energy along z axis) using VNL from the nc file? I already did the DeviceLDOS calculation in the former script..Thank you very much.

[kstokbro]

You can use the example: www.quantumwise.com/documents/tutorials/latest/MolecularDevice/index.html/ for plotting average LDOS along z

[yang su]

Thank you..I just checked the log file of this simulation and it says the residue is in order of 1e14 and the tolerance is 1e-12..It seems it did not converge...What can I do to make the simulation converge better? Thank you very much.

[kstokbro]

I am not sure which part of the calculation this is related to,
can you post your input and output files, then I will take a look