Ising Model

Implement in CUDA the evolution of an Ising model in two dimensions for a given number of steps k. For more information about the Ising model visit the above link. To execute the produced executable files, an Nvidia GPU is necessary. Parameters tuning for optimal performance was implemented on an Nvidia Tesla P100.

Installation

Clone the repository

git clone https://github.com/vagzikopis/IsingCuda.git

Usage

Local Validation Test (Nvidia GPU prerequisite)

cd src
make
./validate_v1 or ./validate_v2 or ./validate_v3 or ./validate_sequential

Local Execution Time Test (Nvidia GPU prerequisite)

cd src
make
./v1 $n $k or ./v2 $n $k or ./v3 $n $k or ./sequential $n $k

Aristotle University HPC Validation Test

Test whether the CUDA implementations return valid results or not

cd src
make
cd ..
cd batch
sbatch validate_v1.sh or sbatch validate_v2.sh or sbatch validate_v3.sh

Aristotle University Time Test (Nvidia GPU prerequisite)

To edit parameters n(square lattice axis) and k(iterations), edit v1.sh or v2.sh or v3.sh files. Results are returned in slurm file formats. Access results with command tail -f slurm* .

cd src
make
cd ..
cd batch
sbatch v1.sh or sbatch v2.sh or sbatch v3.sh

Contributing

Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.