Conway's Game of Life: Parallel vs Distributed

Benchmarked two architectural approaches to concurrent grid simulation. Identified where shared-memory scales better and where networked pub/sub wins.

• Optimised worker pool design through goroutine count tuning against grid partitioning — reduced synchronisation overhead by minimising state contention on shared memory.

• Replaced point-to-point messaging with pub/sub model in distributed version — decoupled coordination, improved fault tolerance, and quantified when networking overhead exceeds parallelism gains.

• Identified bottlenecks via CPU profiling, visualised trade-offs across grid sizes and worker counts using MATLAB — established shared-memory latency advantage vs networked scalability ceiling.

• Full benchmark report and implementation details on GitHub — detailed analysis, code walkthroughs, and performance graphs in the repository.