Skip to content(if available)orjump to list(if available)

Accelerated Game of Life with CUDA / Triton

coke12

Coincidentally I am also writing a cellular automata simulation. I've blindly given your article to my software architect subagent, who has identified several architectural improvements that it can make and has converted these into tasks to farm out to other subagents. Thanks!

Just a glimpse from building software in 2025.

IncreasePosts

Did you confirm the improvements are actually real?

coke12

The description/assessment of tasks is all plausible, but agreed, some of the execution can be surprisingly boneheaded :)

Case in point, I am building a cellular automata-based physics system, and there is seemingly nothing I can do to affirm that row 0 is "down" and row 255 is "up". The system just cannot grok it on a consistent basis. It has the ability to take screenshots, write unit tests, etc, it's just blind to the kind of intuitive logic we get with our human world model. So the code frequently regresses and gravity starts going in the wrong direction.

petermcneeley

The WebGPU tutorial is also Game of life.

https://codelabs.developers.google.com/your-first-webgpu-app...

It would be interesting to see a perf comparision. WebGPU does not have a 64 bit type but it could be emulated with a Vec2u load/store.

pjmlp

CUDA supports more than just C or C++, is exactly part of its appeal.