Publications

Hill, Aaron J.; Donaldson, Jonathon W.; Rothganger, Fredrick H.; Vineyard, Craig M.; Follett, David R.; Follett, Pamela L.; Smith, Michael R.; Verzi, Stephen J.; Severa, William M.; Wang, Felix W.; Aimone, James B.; Naegle, John H.; James, Conrad D.

Abstract not provided.

2017 IEEE International Conference on Rebooting Computing, ICRC 2017 - Proceedings

Hill, Aaron J.; Donaldson, Jonathon W.; Rothganger, Fredrick R.; Vineyard, Craig M.; Follett, David R.; Follett, Pamela L.; Smith, Michael R.; Verzi, Stephen J.; Severa, William M.; Wang, Felix W.; Aimone, James B.; Naegle, John H.; James, Conrad D.

Unlike general purpose computer architectures that are comprised of complex processor cores and sequential computation, the brain is innately parallel and contains highly complex connections between computational units (neurons). Key to the architecture of the brain is a functionality enabled by the combined effect of spiking communication and sparse connectivity with unique variable efficacies and temporal latencies. Utilizing these neuroscience principles, we have developed the Spiking Temporal Processing Unit (STPU) architecture which is well-suited for areas such as pattern recognition and natural language processing. In this paper, we formally describe the STPU, implement the STPU on a field programmable gate array, and show measured performance data.

Proceedings of the International Joint Conference on Neural Networks

Smith, Michael R.; Hill, Aaron J.; Carlson, Kristofor D.; Vineyard, Craig M.; Donaldson, Jonathon W.; Follett, David R.; Follett, Pamela L.; Naegle, John H.; James, Conrad D.; Aimone, James B.

Information in neural networks is represented as weighted connections, or synapses, between neurons. This poses a problem as the primary computational bottleneck for neural networks is the vector-matrix multiply when inputs are multiplied by the neural network weights. Conventional processing architectures are not well suited for simulating neural networks, often requiring large amounts of energy and time. Additionally, synapses in biological neural networks are not binary connections, but exhibit a nonlinear response function as neurotransmitters are emitted and diffuse between neurons. Inspired by neuroscience principles, we present a digital neuromorphic architecture, the Spiking Temporal Processing Unit (STPU), capable of modeling arbitrary complex synaptic response functions without requiring additional hardware components. We consider the paradigm of spiking neurons with temporally coded information as opposed to non-spiking rate coded neurons used in most neural networks. In this paradigm we examine liquid state machines applied to speech recognition and show how a liquid state machine with temporal dynamics maps onto the STPU - demonstrating the flexibility and efficiency of the STPU for instantiating neural algorithms.

Smith, Michael R.; Hill, Aaron J.; Carlson, Kristofor D.; Vineyard, Craig M.; Donaldson, Jonathon W.; Follett, David R.; Follett, Pamela L.; Naegle, John H.; James, Conrad D.; Aimone, James B.; Smith, Michael R.; Smith, Michael R.

Abstract not provided.

Smith, Michael R.; Hill, Aaron J.; Carlson, Kristofor D.; Vineyard, Craig M.; Donaldson, Jonathon W.; Follett, David R.; Follett, Pamela L.; Naegle, John H.; James, Conrad D.; Aimone, James B.

Abstract not provided.

Vineyard, Craig M.; Aimone, James B.; Verzi, Stephen J.; Donaldson, Jonathon W.; Smith, Michael R.; Rothganger, Fredrick R.; Follet, David R.; James, Conrad D.; Naegle, John H.

Abstract not provided.

Smith, Michael R.; Hill, Aaron J.; Carlson, Kristofor D.; Vineyard, Craig M.; Donaldson, Jonathon W.; Follett, David R.; Follett, Pamela L.; Naegle, John H.; James, Conrad D.; Aimone, James B.

Abstract not provided.

Rothganger, Fredrick R.; Follett, David R.; Naegle, John H.; Wang, Felix W.; Donaldson, Jonathon W.; Vineyard, Craig M.; James, Conrad D.; Aimone, James B.

Abstract not provided.