arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2408.12836cs.AR
收藏

面向CNN的近似乘法器的架构错误度量

An Architectural Error Metric for CNN-Oriented Approximate Multipliers

Ao Liu, Jie Han, Qin Wang, Zhigang Mao, Honglan Jiang

AI总结:

本研究分析了AM-CNN中的错误传播,提出架构平均错误(AME)度量评估精度下降,实验证明其优于其他度量,基于AME的二次回归模型能高效预测AM-CNN精度,比GPU仿真快约$10^{6}\ imes$。

摘要:

作为实现卷积神经网络(CNNs)中大量乘法运算的潜在替代方案,近似乘法器(AMs)兼具高硬件效率和高精度的潜力。然而,对精度的表征和设计合适的AM对于基于AM的CNN(AM-CNN)至关重要。在本研究中,通过考虑CNN架构来分析AM-CNN中错误的产生和传播。基于此分析,提出了一种新的AM错误度量来评估AM-CNN的精度下降,称为架构平均错误(AME)。在CIFAR-10、CIFAR-100和ImageNet数据集上的VGG和ResNet中评估了AME的有效性。实验结果表明,AME与AM-CNNs的精度表现出强相关性,优于其他AM错误度量。为了预测AM-CNNs的精度,基于AME构建了二次回归模型;预测结果与真实值的平均偏差为3%。与基于GPU的仿真相比,基于AME的预测快约$10^{6}\ imes$。

英文摘要:

As a potential alternative for implementing the large number of multiplications in convolutional neural networks (CNNs), approximate multipliers (AMs) promise both high hardware efficiency and accuracy. However, the characterization of accuracy and design of appropriate AMs are critical to an AM-based CNN (AM-CNN). In this work, the generation and propagation of errors in an AM-CNN are analyzed by considering the CNN architecture. Based on this analysis, a novel AM error metric is proposed to evaluate the accuracy degradation of an AM-CNN, denoted as the architectural mean error (AME). The effectiveness of the AME is assessed in VGG and ResNet on CIFAR-10, CIFAR-100, and ImageNet datasets. Experimental results show that AME exhibits a strong correlation with the accuracy of AM-CNNs, outperforming the other AM error metrics. To predict the accuracy of AM-CNNs, quadratic regression models are constructed based on the AME; the predictions show an average of 3% deviation from the ground-truth values. Compared with a GPU-based simulation, the AME-based prediction is about $10^{6}\times$ faster.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑