PASS: An Asynchronous Probabilistic Processor for Next Generation Intelligence

TL;DR

PASS is a novel asynchronous probabilistic accelerator built with 14nm CMOS technology, demonstrating significant energy efficiency and broad applicability in solving complex probabilistic problems like optimization, neural simulation, and machine learning.

Contribution

It introduces PASS, the first fully integrated asynchronous probabilistic accelerator leveraging Ising Model parallelism, with extensive demonstrations of its versatility and efficiency.

Findings

Up to 23,000x energy efficiency improvement over CPUs.

Broad applicability across optimization, neural simulation, and machine learning.

First on-chip integrated asynchronous probabilistic accelerator.

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 $23,000$x energy to solution improvement compared to CPUs on probabilistic problems.