Graph Compression, Neural Networks, and More – Using Oscillators to Compute

Abstract: The rapid advancement of digital electronic logic has significantly improved computing capabilities over the past few decades. However, continuing this progress by scaling down devices has become increasingly difficult and costly. This has led to a renewal of interest in alternative materials, devices, and computing strategies.

I present an overview of one promising computing strategy which utilizes oscillatory elements to express and compute with information represented via phases. These systems can compute numerous functions such as state machines, neural networks, and graph compression via novel hardware elements. This provides a strategy for emerging hardware devices with oscillatory properties to be integrated into computing systems to accelerate or improve the efficiency of certain tasks.