Parallel Genetic Algorithm for 0/1 knapsack problem

Summary of Work

• Implemented Parallel Genetic Algorithms using CUDA, OMP, Pthread, MPI, MPI+OMP.
• Addressed challenges like random function overhead efficiently.

Key Findings

• CUDA achieved the highest speedup (~23x), excelling for small-medium workloads.
• CPU solutions provided stable scalability but faced overhead at larger scales.
• Linear execution time observed with ITEMS_NUM and POP_SIZE.

for more detail please reference our presentation and demo video