Berlin 2024 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
DY: Fachverband Dynamik und Statistische Physik
DY 41: Statistical Physics: General
DY 41.8: Talk
Thursday, March 21, 2024, 11:30–11:45, BH-N 128
Analysis of the effects of the entropy source on Monte Carlo simulations — •Anton Lebedev1, Olha Ivanyshyn Yaman1, Annika Möslein2, Zhanet Zaharieva2, and Charles Bryant2 — 1UKRI-STFC Hartree Centre, Keckwick Ln, Warrington, United Kingdom — 2Quantum Dice Ltd, Oxford Centre for Innovation, Oxford, United Kingdom
In this contribution we present the benefits of quantum random number generators (QRNGs) for Monte Carlo simulations using select examples from mathematics and physics. We further present the set of statistical tests performed to arrive at this conclusion when comparing QRNGs to (industry-standard) pseudo-random and radio-based random number generators.
From simple Pi estimation to Bayesian model fitting: Monte Carlo applications are ubiquitous. All rely on randomness to sample the solution space, yet analysis of the quality of random number generators is limited. Understanding the effects of the randomness source on MC simulations and leveraging verifiable quantum randomness will yield a reasonable reduction in the number of simulations required to achieve a prescribed uncertainty bound and thereby the amount of compute resources consumed.
Using examples from mathematics and physics, we have analysed the statistically significant differences in simulation outcomes between quantum RNGs, classical hardware RNGs and (parallel) pseudo-random number generators. Tests for statistical significance have been selected based on the amount of underlying assumptions.
Keywords: Qunatum Random Number Generator; Monte Carlo; Entropy; Statistical tests