Skip to content
High Performance Computing of Fiber Scattering Simulation
Cuda C++ C Shell Makefile
Branch: master
Clone or download

Latest commit

Fetching latest commit…
Cannot retrieve the latest commit at this time.

Files

Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
CPU
CPU_OpenMP
CUDA
MPI_OpenMP
data
.gitattributes
.gitignore
LICENSE
README.md

README.md

FiberScattering

Introduction

The simulation algorithm is designed by Zhang Yan.
We developed optimized single GPU and multi-GPU solutions for this simulation.
Here is a list of different versions we implemented.

  • CPU : single-threaded
  • CPU_OpenMP : parallelize for loop
  • CUDA : GPU implementation for a single GPU system
  • MPI_CUDA : distributed GPU implementation
  • MPI_OpenMP : distributed CPU implementation

Usage

Usage: ./fiber_(cpu/omp/cuda/mpi) [options] -f filename

    -f filename      :file containing atom info
    -l lamda         :angstrom value                 [default=1.033]
    -d distance      :specimen to detector distance  [default=300]
    -s span          :sampling resolution            [default=2048]

Reference / Citation

Please refer to it if you want to use.

@inproceedings{yu2015high,
  title={High performance computing of fiber scattering simulation},
  author={Yu, Leiming and Zhang, Yan and Gong, Xiang and Roy, Nilay and Makowski, Lee and Kaeli, David},
  booktitle={Proceedings of the 8th Workshop on General Purpose Processing using GPUs},
  pages={90--98},
  year={2015},
  organization={ACM}
}
You can’t perform that action at this time.