I am an electrical engineer in the Quantum Nanophotonics and Faint Photonics Group at the National Institute of Standards and Technology, Boulder, CO. I received a PhD in the Lightwave Communications Research Laboratory, Department of Electrical Engineering, Princeton University, Princeton, NJ under the direction of Paul Prucnal.
My home field is physics and integrated photonics. I research specialized computers that employ light in place of electricity. In short,
- Optical pulses can transport information very well – think 100 picosecond action potentials
- Optical devices can’t meet the requirements of digital gates
- It turns out though that some optical devices can meet the requirements of analog neurons, leading to neuromorphic hardware that works at roughly 10^6 of biological time.
Some prospective uses of ultrafast neuromorphic hardware (widely varying in speculativeness) include
- Functionalize with deep frameworks, and classify pictures of cats and pedestrians within a datacenter (startup, startup, startup), or
- Functionalize with recurrent frameworks (such as NEF), and solve time-critical math problems, such as found in quantum state reconstruction, aeronautic control, particle physics, dynamical system emulation, cognitive radio, and trans-oceanic communication, or
- Explore some technological limits of artificial cognition.