This project was intended to introduce me to CUDA programming. I have implemented Conway's Game of Life for both the CPU and GPU, using SFML to provide a GUI.
- a CUDA-capable GPU
- SFML for graphics
- CUDA libraries
- CMake
- build by running
cmake --build .in the root of the repo - to run GUI, run
./run_game - to execute the speed test, run
./speed_test - execute tests by running
ctestin root of directory
This is not a fair comparison between CPU and GPU because there are many optimisations one could introduce to my CPU implemention (e.g. multithreading). However, the 200x run time difference was useful in demonstrating to me the power of a GPU for embarassingly-parallel problems.
