BKS parametrization of buck/coul/long pair_style for silica

Dear All,

My graduate research project focuses on MD simulations of amorphous
silica. I have performed melt-quench MD simulations using LAMMPS with
the tabulated form of the BKS interatomic potential for silica, posted
to the LAMMPS list previously at
http://lammps.sandia.gov/threads/msg22276.html

I am currently using the GPU acceleration package USER-CUDA. In order
to move forward with GPU accelerated simulations I need the SiO2 BKS
parametrization of the buck/coul/long pair_style, to use with
"pair_style buck/coul/long/cuda." Though I have attempted to use the
coefficients listed in the BKS publication [PRL 64, 1955 (1990)] I
have not succeeded in replicating the table values.

I would greatly appreciate if anyone could provide the coefficients
used to produce the tables in the link above. Any other tips or
information that might be offered would be very welcomed.

Thank you in advance for your time and consideration.

Sincerely,

Benjamin N. Doblack
Graduate Student
School of Engineering
University of California at Merced
[email protected]

Dear All,

My graduate research project focuses on MD simulations of amorphous
silica. I have performed melt-quench MD simulations using LAMMPS with
the tabulated form of the BKS interatomic potential for silica, posted
to the LAMMPS list previously at
http://lammps.sandia.gov/threads/msg22276.html

I am currently using the GPU acceleration package USER-CUDA. In order
to move forward with GPU accelerated simulations I need the SiO2 BKS
parametrization of the buck/coul/long pair_style, to use with
"pair_style buck/coul/long/cuda." Though I have attempted to use the
coefficients listed in the BKS publication [PRL 64, 1955 (1990)] I
have not succeeded in replicating the table values.

why don't you just use paul's tables and input template
and use the pair_style table/gpu ?

axel.

Dear Axel,

Thank you for your suggestion.

I have previously successfully run the simulation with the gpu library
provided with LAMMPS, using pair_style table/gpu, however, I was 1)
seeking to understand the "fusion problem," 2) hoping to use the
USER-CUDA library as it seems to perform better in our installation.

Sincerely,

Benjamin N. Doblack
Graduate Student
School of Engineering
University of California at Merced
[email protected]

Dear Axel,

Thank you for your suggestion.

I have previously successfully run the simulation with the gpu library
provided with LAMMPS, using pair_style table/gpu, however, I was 1)
seeking to understand the "fusion problem," 2) hoping to use the
USER-CUDA library as it seems to perform better in our installation.

with respect to point 2)
what is your specific setup and how do you run LAMMPS on it?
it would be extremely helpful for us to know more about when
either of the two GPU accelerated package performs better.
have you been running with one, two or more MPI tasks per GPU?
what kind of GPU and host hardware do you have?

thanks,
    axel.