This week we take a short break from 3D programming topics and go into gaming! Well, sort of…
A few weeks ago I published on my GitHub page a CUDA implementation of Conway’s Game of Life. The code is pretty simple, well in tune with the simplicity of the game.
The implementation can be found here: https://github.com/alesegovia/game-of-life.
If you are not familiar with the game, Conway’s Game of Life is a 0-player game where cells live and die on an infinite 2D grid. The life/death rules are the following, according to Wikipedia:
Every cell interacts with its eight neighbours, which are the cells that are horizontally, vertically, or diagonally adjacent. At each step in time, the following transitions occur:
- Any live cell with fewer than two live neighbours dies, as if caused by under-population.
- Any live cell with two or three live neighbours lives on to the next generation.
- Any live cell with more than three live neighbours dies, as if by overcrowding.
- Any dead cell with exactly three live neighbours becomes a live cell, as if by reproduction.
Conway’s game is excellent for implementing on a GPU, as it involves analyzing the cells in the 2D grid and, what’s best, each cell’s next state depends only on the previous state of its neighbors and never on the their current state.
This means that we can spawn a GPU thread for every single cell in the board and calculate the next state in parallel.
In the published implementation, the board size is 64×64 cells, so we are effectively spawning 4,096 GPU threads to solve every iteration. We do this for one million generations.
The project has been released under a GPLv3 license, so feel free to download, build it, run it, modify it and share it with others under its terms.
If you are looking for a fun weekend project, the game could definitely use an UI. I’ll give you extra points if you can draw it using OpenGL without ever having to copy the board back from GPU memory into system memory ;-)
Enjoy!