A Study with MIC-GPU-CPU Heterogeneous Architecture in Electrocardiograms Simulation of Electrocardiogram Based Whole-Heart Model

In order to study the performance of MIC-GPU-CPU hybrid computing platform, this paper conduct a whole heart simulation model on this platform. And to corroborate the advantages of this heterogeneous architecture environment, we parallelize the electrocardiograms simulation algorithm. Electrocardiograms simulation algorithm is a recommendable case to study the ability of MIC-GPU-CPU architecture for its great deal of computation. The paper addresses the heart modeling program performs in the MIC-GPU-CPU architecture, such architecture performs well in electrocardiograms simulation application, and the bottleneck in this platform with a set of parameter-various benchmarks. The heart simulation results on CPU, MIC, GPU, and MIC-GPU-CPU are exhibited at last.