S. Churcher

About

S. Churcher is a pioneering figure in analog VLSI neural networks, with a focus on pulse-stream signaling techniques for hardware neural computation. Their major contributions center on the design and fabrication of the EPSILON neural network chipset, a large-scale working VLSI device that demonstrated the viability of pulse-stream methods for high-speed, analog neural processing. The EPSILON chip achieved an impressive 360 million synaptic connections per second, marking a significant milestone in hardware neural network performance. Churcher also advanced dynamic weight storage techniques and explored nonvolatile alternatives, including the development of an amorphous silicon memory for neural network applications. Their foundational work, particularly the 1994 paper "Pulse stream VLSI neural networks" (22 citations) and the 1993 paper detailing the EPSILON chipset (13 citations), has influenced subsequent research in neuromorphic engineering and analog computing. Churcher’s contributions remain relevant for students and researchers interested in efficient, hardware-based neural computation and the historical evolution of VLSI neural systems.

Research Focus

Key Achievements

Top Papers

1. 1
   Pulse stream VLSI neural networks

   22 citations · 1994
2. 2
   PULSE STREAM VLSI CIRCUITS AND SYSTEMS: THE EPSILON NEURAL NETWORK CHIPSET

   13 citations · 1993

Key Collaborators

Contact & Links