提出FAMES方法，在2位量化模型中高效替换乘法器，大幅节能且不损失精度。
FAMES: Fast Approximate Multiplier Substitution for Mixed-Precision Quantized DNNs--Down to 2 Bits!
- 设计快速替代方案，将近似乘法器用于2位混合精度模型
- 平均节能28.67%，精度损失低于1%
- 比传统遗传算法快300倍，适合低比特推理加速场景
在设计能效高的深度神经网络（DNN）加速器时，量化是一种常用技术。近期进展已将DNN的位宽降低至2位。与此同时，许多工作采用近似乘法器（AppMul）以降低能耗。然而，这些方法仍假设位宽远高于2位，与当前量化前沿严重脱节，甚至质疑了在低比特下使用近似乘法器的意义——因为高比特近似乘法器能耗远高于低比特精确乘法器！因此，核心问题是：近似乘法器能否有效应用于极低比特的量化模型？本文给出肯定回答，并提出系统性解决方案：FAMES，一种针对混合精度量化DNN的快速近似乘法器替换方法。实验表明，该方法在先进混合精度量化模型上实现平均28.67%的能耗降低，精度损失控制在1%以内；且相比以往基于遗传算法的方法，速度提升高达300倍。
原文摘要 · Abstract (English)
A widely-used technique in designing energy-efficient deep neural network (DNN) accelerators is quantization. Recent progress in this direction has reduced the bitwidths used in DNN down to 2. Meanwhile, many prior works apply approximate multipliers (AppMuls) in designing DNN accelerators to lower their energy consumption. Unfortunately, these works still assume a bitwidth much larger than 2, which falls far behind the state-of-the-art in quantization area and even challenges the meaningfulness of applying AppMuls in DNN accelerators, since a high-bitwidth AppMul consumes much more energy than a low-bitwidth exact multiplier! Thus, an important problem to study is: Can approximate multipliers be effectively applied to quantized DNN models with very low bitwidths? In this work, we give an affirmative answer to this question and present a systematic solution that achieves the answer: FAMES, a fast approximate multiplier substitution method for mixed-precision DNNs. Our experiments demonstrate an average 28.67% energy reduction on state-of-the-art mixed-precision quantized models with bitwidths as low as 2 bits and accuracy losses kept under 1%. Additionally, our approach is up to 300x faster than previous genetic algorithm-based methods.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。