Mathematical modeling of the electrical activity in cardiac cells is computationally challenging because differential equations describing current flow must be solved at high spatial and temporal resolution. The ionic currents that determine the transmembrane potential are regulated by the solution of time and voltage dependent gating variable equations. In considering an alternate computing strategy to the more traditional serial implementation, we compared the performance of parallel implementations using a prototype model whose equations were defined by the membrane kinetics. Parallel computation was controlled by a `master' process that distributed time step and transmembrane potential information to a variable number of `slave' processes. We tested two levels of parallelization. In `gate-level parallelization', individual gating variables were integrated numerically on six slaves and passed back to the master to complete computation of the currents and transmembrane potential. In `current-level parallelization', ionic currents incorporating gating variable calculations were determined by only four slaves and returned to the master. The computation of gate-level parallelization spread over six CPUs increased wall clock time by 25% compared to the serial case, while the increased workload by the slaves and reduced communication across CPUs in the current-level parallelization decreased wall clock time by 9%.