The high-profile applications of machine learning (ML)/AI, while impressive, are a) not suitable for Size, Weight, and Power (SWaP) limited systems and b) not operable without access to “the cloud” for remote data processing.  Neuromorphic computing [1] is one of the most promising approaches for low-power, non-cloud-tethered ML, also called edge computing [2].  Since neuromorphic computing emulates aspects of biological brains, e.g., trainable networks of neurons and synapses, in non-traditional, highly parallelizable, reconfigurable hardware, it could be much more efficient and adaptable than typical ML approaches.  Ideally, neuromorphic computing should leverage “the physics of the device” to perform the computations and for the reconfigurable hardware itself to be the ML algorithm. This research effort encompasses mathematical models, hardware characterization, hardware emulation, hybrid CMOS architecture designs, and algorithm development for neuromorphic computing processors. We are particularly interested in approaches that exploit the characteristic behavior of the physical hardware itself to perform computation, e.g., optics/photonics, memristors/ReRAM, metamaterials, nanowires, superconductors. Again, special emphasis will be placed on imaginative technologies and solutions to satisfy future Air Force and Space Force needs for non-cloud-tethered ML on SWaP limited assets.

 

[1] Schuman, C.D., Kulkarni, S.R., Parsa, M. et al. Opportunities for neuromorphic computing algorithms and applications. Nat Comput Sci 2, 10–19 (2022).

[2] K. Cao, Y. Liu, G. Meng and Q. Sun, "An Overview on Edge Computing Research," in IEEE Access, vol. 8, pp. 85714-85728, 2020

key words

neuromorphic computing; machine learning; artificial intelligence; edge computing; neuromorphic hardware

Citizenship:  Open to U.S. citizens

Level:  Open to Postdoctoral and Senior applicants