Parallel bin packing using first fit and k-delayed best-fit heuristics

The authors describe the development of parallel implementations for the bin packing problem. Bin packing algorithms are studied to understand various resource allocation issues and the impact of the different packing heuristics on the packing efficiency. The seemingly serial nature of the bin packing simulation has prohibited previous experimentations from going beyond sizes of several thousands bins. The authors show that by adopting fairly simple data parallel algorithms a linear speedup is possible. Sizes of up to hundreds of thousands of bins have been simulated for different parameters and heuristics. The authors focussed on the well known first-fit heuristic. They also considered another potentially superior heuristic, k-delayed best fit.>