Keynote Speakers

This presentation will chart the rise of SPAD devices from their first integration in CMOS technologies to volume production in smartphones and ranging products towards emerging applications in automotive LIDAR and healthcare. This new generation of quantum consumer imaging devices has been enabled by the co-integration of large SPAD arrays, massively parallel photon timing and counting electronics and embedded signal and image processing. The full realisation of the potential of these detectors poses three significant challenges for technologists and circuit and system designers; achieving simultaneously (1) maximal effective quantum efficiency (2) picosecond timing resolution at low power consumption and (3) high temporal-event dynamic range. This talk will be illustrated with pixel circuit solutions and video images drawn from example applications from microscopy, endoscopy, proximity detection, time of flight 3D imaging to flash LIDAR. The opportunities for circuit innovation offered by recent 3D integration of backside illuminated SPADs above nanoscale CMOS will be highlighted as well as trends and prospects for the future.

Massively parallel, intracellular recording of a large number of mammalian neurons across a network has been a great technological pursuit in neurobiology, but it has not been achieved until our recent breakthrough [1]. For example, the intracellular recording by the patch clamp revolutionized neurobiology with its unparalleled sensitivity that can measure down to subthreshold synaptic activities, but it is too bulky to scale into a dense array, and only ~10 parallel patch recordings have so far been possible. For another example, the microelectrode array (MEA) can record from many more neurons, but this extracellular technique has too low a sensitivity to tap into synaptic events. In this talk, I will share the recent breakthrough of ours [1], a CMOS nanoelectrode array that massively parallelizes the intracellular recording from thousands of connected mammalian neurons. I will also explore the applications of this unprecedented tool in fundamental and applied neurobiology, in particular, functional connectome mapping, high-throughput drug screening for neurological disorder, and copying biological neuronal networks as a possible new synthesis of machine intelligence.

[1] J. Abbott et al, “A nanoelectrode array for obtaining intracellular recordings from thousands of connected neurons,”  Nature Biomed. Eng., doi: 10.1038/s41551-019-0455-7 (2019)

Analog computation has a long history, but bit-based digital computers have proven to be superior for general-purpose computation. As main-stream digital technology faces technological and even fundamental limits, especially for power dissipation, alternative analog approaches to computing have recently received increased attention. In such analog dynamical systems, the computational process more closely exploits the underlying physics-driven dynamics, which offers the promise of lower power dissipation. We will review recent work in this area.  Specifically, we will consider physical processes and dynamical systems based on waves and on coupled oscillators.  It is unlikely that such dynamical systems will find use as general-purpose computers, but they might well complement and enhance a digital computing engine for certain tasks that naturally map onto the underlying physics.