Grid World in CUDA

This repo contains the code that was written to evaluate a cuda implementation of the markov decision process.
More specifically the use of value itteration to calculate the optimal policy for navigating a grid world.

This repo is broken up into the following sections

Cases

This is where all the different cases that were tested are stored.
The files contain the utility for each square in a grid world.
This folder also contains the script used to generate worlds of any dimension.

python

This contains a python implementation of value itteration. As it is python, it is quite slow. It's main purpose is to compare the results as it's a known good solution to the grid world problem.

c_serial

This contains a serial implementation written in C for using in speedup comparisons.

cuda

This contains the cuda implementation.