Approximate Multiplier Induced Error Propagation in Deep Neural Networks

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Alahakoon, A. M. H. H., Saadat, Hassaan, Jayasinghe, Darshana, Parameswaran, Sri
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866914185089122304
author Alahakoon, A. M. H. H.
Saadat, Hassaan
Jayasinghe, Darshana
Parameswaran, Sri
author_facet Alahakoon, A. M. H. H.
Saadat, Hassaan
Jayasinghe, Darshana
Parameswaran, Sri
contents Deep Neural Networks (DNNs) rely heavily on dense arithmetic operations, motivating the use of Approximate Multipliers (AxMs) to reduce energy consumption in hardware accelerators. However, a rigorous mathematical characterization of how AxMs error distributions influence DNN accuracy remains underdeveloped. This work presents an analytical framework that connects the statistical error moments of an AxM to the induced distortion in General Matrix Multiplication (GEMM). Using the Frobenius norm of the resulting error matrix, we derive a closed form expression for practical DNN dimensions that demonstrates the distortion is predominantly governed by the multiplier mean error (bias). To evaluate this model in realistic settings, we incorporate controlled error injection into GEMM and convolution layers and examine its effect on ImageNet scale networks. The predicted distortion correlates strongly with the observed accuracy degradation, and an error configurable AxM case study implemented on an FPGA further confirms the analytical trends. By providing a lightweight alternative to behavioral or hardware level simulations, this framework enables rapid estimation of AxM impact on DNN inference quality.
format Preprint
id arxiv_https___arxiv_org_abs_2512_06537
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Approximate Multiplier Induced Error Propagation in Deep Neural Networks
Alahakoon, A. M. H. H.
Saadat, Hassaan
Jayasinghe, Darshana
Parameswaran, Sri
Hardware Architecture
Machine Learning
Deep Neural Networks (DNNs) rely heavily on dense arithmetic operations, motivating the use of Approximate Multipliers (AxMs) to reduce energy consumption in hardware accelerators. However, a rigorous mathematical characterization of how AxMs error distributions influence DNN accuracy remains underdeveloped. This work presents an analytical framework that connects the statistical error moments of an AxM to the induced distortion in General Matrix Multiplication (GEMM). Using the Frobenius norm of the resulting error matrix, we derive a closed form expression for practical DNN dimensions that demonstrates the distortion is predominantly governed by the multiplier mean error (bias). To evaluate this model in realistic settings, we incorporate controlled error injection into GEMM and convolution layers and examine its effect on ImageNet scale networks. The predicted distortion correlates strongly with the observed accuracy degradation, and an error configurable AxM case study implemented on an FPGA further confirms the analytical trends. By providing a lightweight alternative to behavioral or hardware level simulations, this framework enables rapid estimation of AxM impact on DNN inference quality.
title Approximate Multiplier Induced Error Propagation in Deep Neural Networks
topic Hardware Architecture
Machine Learning
url https://arxiv.org/abs/2512.06537