As the complexity of neural models continues to increase (larger populations, varied ionic conductances, more detailed morphologies, etc) traditional software-based models have difficulty scaling to reach the performance levels desired. This paper describes the use of FPGAs, or field programmable gate arrays, to easily implement a wide variety of neural models with the performance of custom analogue circuits or computer clusters, the reconfigurability of software, and at a cost rivalling personal computers. FPGAs reach this level of performance by enabling the design of neural models as parallel processed data paths. These architectures provide for a wide range of single-compartment, multi-compartment and population models to be readily converted to FPGA implementations. Generalized architectures are described for the efficient modelling of a first-order, nonlinear differential equation in throughput maximizing or latency minimizing data-path configurations. The homogeneity of population and multicompartment models is exploited to form deep pipelines for improved performance. Limitations of FPGA architectures and future research areas are explored.

Field programmable gate arrays (FPGAs) have previously been shown as high-performance platforms for neural-modeling applications. Implementations have traditionally been time-consuming and error-prone, requiring the neural modeler to work outside of their expert domain. This paper demonstrates a new approach to the development of neural models using an auto-generation toolkit. This design tool enables model construction-level alterations (e.g., adjustment of model population size or insertion/deletion of an ionic conductance) within hours and parameter changes on-the-fly. The approach is validated on a 40-neuron pre-Bötzinger complex population model consisting of Hodgkin-Huxley style conductances and fully interconnected synapses. A total of 1880 parameters are on-the-fly user tunable on a free-running model. The resulting implemented model performs at a theoretical 8.7 x real-time utilizing 90% of logic elements within a Xilinx Virtex-4 XC4VSX35-fg676-10 FPGA.

Physiologically realistic models of motoneurons are often simulated using traditional software modeling tools, with suboptimal performance. This paper describes the successful implementation of a multi-compartment motoneuron model demonstrating plateau potentials on a Field Programmable Gate Array (FPGA). Simulation performance exceeds 4x realtime, an over 12x improvement over a software simulation. This demonstrates the capability of FPGAs in executing complex dynamical systems with high throughput.

Numerical simulations of dynamical systems are an obvious application of high-performance computing. Unfortunately, this application is underutilized because many modelers lack the technical expertise and financial resources to leverage high-performance computing hardware. Additionally, few platforms exist that can enable high-performance computing with real-time guarantees for inclusion into embedded systems--a prerequisite for working with medical devices. Here we introduce simEngine, a platform for numerical simulations of dynamical systems that reduces modelers' programming effort, delivers simulation speeds 10-100 times faster than a conventional microprocessor, and targets high-performance hardware suitable for real-time and embedded applications. This platform consists of a high-level mathematical language used to describe the simulation, a compiler/resource scheduler that generates the high-performance implementation of the simulation, and the high-performance hardware target. In this paper we present an overview of the platform, including a network-attached embedded computing device utilizing field-programmable gate arrays (FPGAs) suitable for real-time, high-performance computing. We go on to describe an example model implementation to demonstrate the platform's performance and describe how future development will improve system performance.