Numerical computation is essential to many areas of artificial intelligence (AI), whose computing demands continue to grow dramatically, yet their continued scaling is jeopardized by the slowdown in Moore's law. Multi-function multi-way analog (MFMWA) technology, a computing architecture comprising arrays of memristors supporting in-memory computation of matrix operations, can offer tremendous improvements in computation and energy, but at the expense of inherent unpredictability and noise. We devise novel randomized algorithms tailored to MFMWA architectures that mitigate the detrimental impact of imperfect analog computations while realizing their potential benefits across various areas of AI, such as applications in computer vision. Through analysis, measurements from analog devices, and simulations of larger systems, we demonstrate orders of magnitude reduction in both computation and energy with accuracy similar to digital computers.
Cite
@article{arxiv.2401.13754,
title = {Multi-Function Multi-Way Analog Technology for Sustainable Machine Intelligence Computation},
author = {Vassilis Kalantzis and Mark S. Squillante and Shashanka Ubaru and Tayfun Gokmen and Chai Wah Wu and Anshul Gupta and Haim Avron and Tomasz Nowicki and Malte Rasch and Murat Onen and Vanessa Lopez Marrero and Effendi Leobandung and Yasuteru Kohda and Wilfried Haensch and Lior Horesh},
journal= {arXiv preprint arXiv:2401.13754},
year = {2024}
}