Benchmarking a Probabilistic Coprocessor

TL;DR

This paper introduces a probabilistic coprocessor based on p-bits, designed to efficiently generate random numbers for Monte Carlo algorithms, and demonstrates its potential to outperform classical CPU and GPU implementations significantly.

Contribution

The paper presents a novel probabilistic coprocessor architecture based on p-bits and benchmarks its performance, showing potential for orders-of-magnitude improvements over classical hardware.

Findings

Probabilistic coprocessor can outperform CPUs and GPUs by multiple orders of magnitude.

P-bits are naturally suited for Monte Carlo algorithms.

Benchmark results demonstrate significant speedups.

Abstract

Computation in the past decades has been driven by deterministic computers based on classical deterministic bits. Recently, alternative computing paradigms and domain-based computing like quantum computing and probabilistic computing have gained traction. While quantum computers based on q-bits utilize quantum effects to advance computation, probabilistic computers based on probabilistic (p-)bits are naturally suited to solve problems that require large amount of random numbers utilized in Monte Carlo and Markov Chain Monte Carlo algorithms. These Monte Carlo techniques are used to solve important problems in the fields of optimization, numerical integration or sampling from probability distributions. However, to efficiently implement Monte Carlo algorithms the generation of random numbers is crucial. In this paper, we present and benchmark a probabilistic coprocessor based on p-bits that are naturally suited to solve these problems. We present multiple examples and project that a nanomagnetic implementation of our probabilistic coprocessor can outperform classical CPU and GPU implementations by multiple orders of magnitude.