| Date |
Topic |
Lecture Materials |
| 9/30, 10/2 |
Biophysical foundations of natural intelligence in neural
systems. Subthreshold MOS silicon models of membrane
excitability. Silicon neurons. Hodgkin-Huxley and
integrate-and-fire models of spiking neuronal dynamics.
Action potentials as address events. |
W1 Lectures |
| 10/7, 10/9 |
Silicon retina. Low-noise, high-dynamic range
photoreceptors. Focal-plane array signal processing.
Spatial and temporal contrast sensitivity and adaptation.
Dynamic vision sensors. |
W2 Lectures |
| 10/14, 10/16 |
Silicon cochlea. Low-noise acoustic sensing and
automatic gain control. Continuous wavelet filter banks.
Interaural time difference and level difference auditory
localization. Blind source separation and independent
component analysis. |
W3 Lectures |
| 10/21, 10/23 |
Silicon cortex. Neural and synaptic compute-in-memory
arrays. Address-event decoders and arbiters, and
integrate-and-fire array transceivers. Hierarchical
address-event routing for locally dense, globally sparse
long-range connectivity across vast spatial scales. |
W4 Lectures |
| 10/28, 10/30 |
Midterm review. Modular and scalable design for neuromorphic and
bioelectronic integrated circuits and systems. Design for
full testability and controllability. |
W5 Lectures |
| 11/4, 11/6 |
Low-noise, low-power design.
Fundamental limits of noise-energy efficiency, and metrics of
performance. Biopotential and electrochemical recording and
stimulation, lab-on-a-chip electrophysiology, and neural
interface systems-on-chip. |
W6 Lectures |
| 11/13 |
Learning and adaptation to compensate for external and
internal variability over extended time scales. Background
blind calibration of device mismatch. Correlated double
sampling and chopping for offset drift and low-frequency
noise cancellation. |
W7 Lectures |
| 11/18, 11/20 |
Adaptive offset cancelation and autoranging in dynamic vision sensing. Tobi Delbruck's lecture on silicon retina history with a live demo of event-based dynamic vision systems. |
W8 Lecture |
11/25, 11/27
|
Energy conservation. Resonant inductive power delivery
and data telemetry. Ultra-high efficiency neuromorphic
computing. Resonant adiabatic energy-recovery
charge-conserving synapse arrays.
|
W9 Lectures |
| 12/2 - 12/6 |
Project final presentations. All are welcome! |
Fall 2024 Projects |
