ASDL: A Fast Vector Approximation Algorithm for Multiplierless Designs of Convolutional Neural Networks

Matrix multiplications are one of the main computational bottlenecks in neural networks. The calculation reduces to numerous sums of partial products, which scale with the total number of Signed Power of Two (SPT) values in the multiplier matrix. Minimizing the SPT count can significantly reduce the computational burden.In this context, we present a Sum Of (signed) Powers Of Two (SOPOT) vector approximation algorithm that limits SPT counts, namely the Accelerated Signed Digit Loading (ASDL) algorithm. This approximation algorithm converts ResNet50 $10$ times faster than its predecessor MPGBP, while being more precise. We apply this method to the pre-trained coefficients of different convolutional neural networks, achieving $80.41\%$ ImageNet accuracy down from $81.3\%$, at an average of only $1.5$ active bits per coefficient. This is done without any retraining or calibration, which are typically required for the most aggressive quantization methods. Our approach offers a promising direction for neural network inference with higher energy efficiency and fewer hardware requirements: we anticipate up to $10$ times improvement over float16 inference.

Authors

Publication Details

Journal
arXiv (Cornell University)
Published
2026-11-16
DOI
https://doi.org/10.48550/arxiv.2611.21103
Primary Topic
Advanced Neural Network Applications
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

ASDL: A Fast Vector Approximation Algorithm for Multiplierless Designs of Convolutional Neural Networks

Luiz F. da S. Coelho, Didier Le Ruyet, Lounis Zerioul, Guillaume H. Tong
arXiv (Cornell University)
Advanced Neural Network Applications
preprint

ASDL: A Fast Vector Approximation Algorithm for Multiplierless Designs of Convolutional Neural Networks

Luiz F. da S. Coelho, Didier Le Ruyet, Lounis Zerioul, Guillaume H. Tong
preprint en

Abstract

Matrix multiplications are one of the main computational bottlenecks in neural networks. The calculation reduces to numerous sums of partial products, which scale with the total number of Signed Power of Two (SPT) values in the multiplier matrix. Minimizing the SPT count can significantly reduce the computational burden.In this context, we present a Sum Of (signed) Powers Of Two (SOPOT) vector approximation algorithm that limits SPT counts, namely the Accelerated Signed Digit Loading (ASDL) algorithm. This approximation algorithm converts ResNet50 $10$ times faster than its predecessor MPGBP, while being more precise. We apply this method to the pre-trained coefficients of different convolutional neural networks, achieving $80.41\%$ ImageNet accuracy down from $81.3\%$, at an average of only $1.5$ active bits per coefficient. This is done without any retraining or calibration, which are typically required for the most aggressive quantization methods. Our approach offers a promising direction for neural network inference with higher energy efficiency and fewer hardware requirements: we anticipate up to $10$ times improvement over float16 inference.

arXiv (Cornell University)
Advanced Neural Network Applications
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon