Distributed computing for membrane-based modeling of action potential propagation

Action potential propagation simulations with physiologic membrane currents and macroscopic tissue dimensions are computationally expensive. We, there fore, analyzed distributed computing schemes to reduce execution time in wo rkstation clusters by parallelizing solutions with message passing. Four sc hemes were considered in two-dimensional monodomain simulations with the Be eler-Reuter membrane equations. Parallel speedups measured with each scheme were compared to theoretical speedups, recognizing the relationship betwee n speedup and code portions that executed serially. A data decomposition sc heme based on total ionic current provided the best performance. Analysis o f communication latencies in that scheme led to a load-balancing algorithm in which measured speedups at 89 +/- 2% and 75 +/- 8% of theoretical speedu ps were achieved in homogeneous and heterogeneous clusters of workstations. Speedups in this scheme with the Luo-Rudy dynamic membrane equations excee ded 3.0 with eight distributed workstations. Cluster speedups were comparab le to those measured during parallel execution on a shared memory machine.