This paper presents a study of the applicability of clusters of GPUs to high-resolution 3D simulations of cardiac electrophysiology. By experimenting with representative cardiac cell models and ODE solvers, in association with solving the monodomain equation, we quantitatively analyze the obtainable computational capacity of GPU clusters. It is found that for a 501×501×101 3D mesh, which entails a 0.1mm spatial resolution, a 128-GPU cluster only needs a few minutes to carry out a 100,000-time-step cardiac excitation simulation that involves a four-variable cell model. Even higher spatial and temporal resolutions are achievable for such simplified mathematical models. On the other hand, our experiments also show that a dramatically larger cluster of GPUs is needed to handle a very detailed cardiac cell model.

Aiming to ease the parallel programming for heterogeneous architectures, we propose and implement a high-level OpenCL runtime that conceptually merges multiple heterogeneous hardware devices into one virtual heterogeneous compute device (VHCD). Moreover, automated workload distribution among the devices is based on offline profiling, together with new programming directives that define the device-independent data access range per work-group. Therefore, an OpenCL program originally written for a single compute device can, after inserting a small number of programming directives, run efficiently on a platform consisting of heterogeneous compute devices. Performance is ensured by introducing the technique of virtual cache management, which minimizes the amount of host-device data transfer. Our new OpenCL runtime is evaluated by a diverse set of OpenCL benchmarks, demonstrating good performance on various configurations of a heterogeneous system.