Unconventional Computing from Qubits to Coupled Oscillators

This talk will cover my research activities in both quantum and neuromorphic computing. The first part of the talk will focus on hardware-software co-design methods to investigate quantum circuit implementation and error mitigation techniques for NISQ computing. I will discuss the qubit mapping problem for quantum circuit implementation in superconducting qubits via IBM QISKIT.  In the second part, I will introduce a novel computing paradigm and architecture which solves hard combinatorial optimization problems (COPs) using the collective dynamics of coupled electrical oscillators as an Ising machine. I will present some of the design of unconventional computing paradigm with oscillatory neural networks that efficiently solves COPs in continuous time with a significant runtime improvement and a low energy footprint compared to digital electronics. Computing with oscillators has shown a significant impact on designing Ising machines to solve NP-hard problems and additionally as an energy efficient hardware accelerator for AI applications. Then, I will conclude with the current challenges and progress in energy efficient computing.