Monte Carlo Simulation for Pi

Contains the following:

1. A serial approach to the estimation of Pi using the Monte Carlo Method (C++)

2. A Multithreaded estimation of Pi based on the Monte Carlo Method (C++)

3. Using matplotlib, plots the estimated value of Pi using Monte Carlo Method (Python)

10000 Iterations

100000 Iterations

10000 Iterations Zoomed

4. A parallel computing version that estimates Pi using the Monte Carlo Method in OpenMP (C)

To use OpenMP

/usr/local/opt/llvm/bin/clang -fopenmp -L/usr/local/opt/llvm/lib main.c -o a

./a

5. A parallel computing version that estimates Pi using the Monte Carlo Method in MPI (C)

To use MPI

The best performance is achieved when the number of nodes is equal to number of cores. E.g.:

$HOME/opt/usr/local/bin/mpicc -o pi ./mpi_pi.c

$HOME/opt/usr/local/bin/mpirun -np 4 ./pi

MPI

MPI Reduction

MPI Send & Receive