PASS: An Asynchronous Probabilistic Processor for Next Generation Intelligence
Abstract
New computing paradigms are required to solve the most challenging computational problems where no exact polynomial time solution exists.Probabilistic Ising Accelerators has gained promise on these problems with the ability to model complex probability distributions and find ground states of intractable problems. In this context, we have demonstrated the Parallel Asynchronous Stochastic Sampler (PASS), the first fully on-chip integrated, asynchronous, probabilistic accelerator that takes advantage of the intrinsic fine-grained parallelism of the Ising Model and built in state of the art 14nm CMOS FinFET technology. We have demonstrated broad applicability of this accelerator on problems ranging from Combinatorial Optimization, Neural Simulation, to Machine Learning along with up to x energy to solution improvement compared to CPUs on probabilistic problems.
Cite
@article{arxiv.2409.10325,
title = {PASS: An Asynchronous Probabilistic Processor for Next Generation Intelligence},
author = {Saavan Patel and Philip Canoza and Adhiraj Datar and Steven Lu and Chirag Garg and Sayeef Salahuddin},
journal= {arXiv preprint arXiv:2409.10325},
year = {2024}
}
Comments
13 page main text, 5 main figures, 21 pages supplementary and methods, 7 supplementary figures, 2 supplementary tables