基于非易失性存储器阵列的Quantization抗噪LSTM神经网络
Non-Volatile Memory Array Based Quantization- and Noise-Resilient LSTM Neural Networks
摘要:
在云计算和边缘计算模型中,边缘计算设备尽可能节能非常重要。长短期记忆(LSTM)神经网络已被广泛用于自然语言处理、时间序列预测和许多其他序列数据任务。因此,对于这些应用,越来越需要低功耗加速器来进行边缘LSTM模型推理。为了减少由于推断设备内的数据传输而引起的功率耗散,已经对使用非易失性存储器(NVM)权重阵列来加速向量矩阵乘法(VMM)操作产生了极大的兴趣。在基于NVM阵列的硬件中,减小的位宽也显著地提高了功率效率。在本文中,我们专注于量化感知训练算法在LSTM模型中的应用,以及这些模型在抵抗量化误差和模拟器件噪声方面带来的好处。我们已经证明,只需要4位NVM权重和4位ADC/DAC就可以产生与浮点基线相当的LSTM网络性能。合理水平的ADC量化噪声和权重噪声可以在我们基于NVM的量化LSTM网络中自然容忍。对我们提出的LSTM推理加速器的基准分析表明,与传统数字方法(GPU,FPGA和ASIC)相比,计算效率至少提高了2.4倍,面积效率提高了40倍。其他一些基于NVM的新方法有望提供更高的计算效率(高达4.7倍),但需要更大的阵列,潜在的更高的出错率。
英文摘要:
In cloud and edge computing models, it is important that compute devices at the edge be as power efficient as possible. Long short-term memory (LSTM) neural networks have been widely used for natural language processing, time series prediction and many other sequential data tasks. Thus, for these applications there is increasing need for low-power accelerators for LSTM model inference at the edge. In order to reduce power dissipation due to data transfers within inference devices, there has been significant interest in accelerating vector-matrix multiplication (VMM) operations using non-volatile memory (NVM) weight arrays. In NVM array-based hardware, reduced bit-widths also significantly increases the power efficiency. In this paper, we focus on the application of quantization-aware training algorithm to LSTM models, and the benefits these models bring in terms of resilience against both quantization error and analog device noise. We have shown that only 4-bit NVM weights and 4-bit ADC/DACs are needed to produce equivalent LSTM network performance as floating-point baseline. Reasonable levels of ADC quantization noise and weight noise can be naturally tolerated within our NVMbased quantized LSTM network. Benchmark analysis of our proposed LSTM accelerator for inference has shown at least 2.4x better computing efficiency and 40x higher area efficiency than traditional digital approaches (GPU, FPGA, and ASIC). Some other novel approaches based on NVM promise to deliver higher computing efficiency (up to 4.7x) but require larger arrays with potential higher error rates.
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。