Exploring Parallelism in FPGA-Based Accelerators for Machine Learning Applications

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Centeno, Sed, Sprague, Christopher, Purkayastha, Arnab A, Simar, Ray, Magotra, Neeraj
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866914158872625152
author Centeno, Sed
Sprague, Christopher
Purkayastha, Arnab A
Simar, Ray
Magotra, Neeraj
author_facet Centeno, Sed
Sprague, Christopher
Purkayastha, Arnab A
Simar, Ray
Magotra, Neeraj
contents Speculative backpropagation has emerged as a promising technique to accelerate the training of neural networks by overlapping the forward and backward passes. Leveraging speculative weight updates when error gradients fall within a specific threshold reduces training time without substantially compromising accuracy. In this work, we implement speculative backpropagation on the MNIST dataset using OpenMP as the parallel programming platform. OpenMP's multi-threading capabilities enable simultaneous execution of forward and speculative backpropagation steps, significantly improving training speed. The application is planned for synthesis on a state-of-the-art FPGA to demonstrate its potential for hardware acceleration. Our CPU-based experimental results demonstrate that speculative backpropagation achieves a maximum speedup of 24% in execution time when using a threshold of 0.25, and accuracy remaining within 3-4% of the baseline across various epochs. Additionally, when comparing individual step execution time, speculative backpropagation yields a maximum speedup of 35% over the baseline, demonstrating the effectiveness of overlapping forward and backward passes.
format Preprint
id arxiv_https___arxiv_org_abs_2511_11640
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Exploring Parallelism in FPGA-Based Accelerators for Machine Learning Applications
Centeno, Sed
Sprague, Christopher
Purkayastha, Arnab A
Simar, Ray
Magotra, Neeraj
Distributed, Parallel, and Cluster Computing
Hardware Architecture
Machine Learning
Speculative backpropagation has emerged as a promising technique to accelerate the training of neural networks by overlapping the forward and backward passes. Leveraging speculative weight updates when error gradients fall within a specific threshold reduces training time without substantially compromising accuracy. In this work, we implement speculative backpropagation on the MNIST dataset using OpenMP as the parallel programming platform. OpenMP's multi-threading capabilities enable simultaneous execution of forward and speculative backpropagation steps, significantly improving training speed. The application is planned for synthesis on a state-of-the-art FPGA to demonstrate its potential for hardware acceleration. Our CPU-based experimental results demonstrate that speculative backpropagation achieves a maximum speedup of 24% in execution time when using a threshold of 0.25, and accuracy remaining within 3-4% of the baseline across various epochs. Additionally, when comparing individual step execution time, speculative backpropagation yields a maximum speedup of 35% over the baseline, demonstrating the effectiveness of overlapping forward and backward passes.
title Exploring Parallelism in FPGA-Based Accelerators for Machine Learning Applications
topic Distributed, Parallel, and Cluster Computing
Hardware Architecture
Machine Learning
url https://arxiv.org/abs/2511.11640