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.

Sandia National Laboratories is using a set of evolving technologies to develop a standards-based approach to wide- and local-area networking, which offers the potential of gigabit speeds. In particular, asynchronous transfer mode (ATM) switches and synchronous optical network (SONET) technologies were used to build a supercomputing network between its California and New Mexico sites and now is being deployed in the local-area environment. The progress of these endeavors and the lessons learned are discussed.>