Analog in-memory computing (AIMC) -- a promising approach for energy-efficient acceleration of deep learning workloads -- computes matrix-vector multiplications (MVMs) but only approximately, due to nonidealities that often are non-deterministic or nonlinear. This can adversely impact the achievable deep neural network (DNN) inference accuracy as compared to a conventional floating point (FP) implementation. While retraining has previously been suggested to improve robustness, prior work has explored only a few DNN topologies, using disparate and overly simplified AIMC hardware models. Here, we use hardware-aware (HWA) training to systematically examine the accuracy of AIMC for multiple common artificial intelligence (AI) workloads across multiple DNN topologies, and investigate sensitivity and robustness to a broad set of nonidealities. By introducing a new and highly realistic AIMC crossbar-model, we improve significantly on earlier retraining approaches. We show that many large-scale DNNs of various topologies, including convolutional neural networks (CNNs), recurrent neural networks (RNNs), and transformers, can in fact be successfully retrained to show iso-accuracy on AIMC. Our results further suggest that AIMC nonidealities that add noise to the inputs or outputs, not the weights, have the largest impact on DNN accuracy, and that RNNs are particularly robust to all nonidealities.
翻译:模拟模拟计算(AIMC) -- -- 以节能加速深层次学习工作量的一种很有希望的方法 -- -- 计算矩阵-矢量倍增(MMMMM),但由于非理想性往往非确定性或非线性,因此只是粗略地计算。这可能会对可实现的深神经网络(DNN)推导准确性产生消极影响,而与常规浮动点(FP)相比,这可能会对可实现的深神经网络(DNN)推导准确性产生消极影响。虽然以前曾建议再培训是为了提高稳健性,但先前的工作只探索了少数DNNNE的表层,使用了不同和过于简化的AIMC硬件模型。在这里,我们使用硬件觉变异和过于简化的培训系统(HWA)系统地审查AIMC对于多种通用人工智能(AI)工作量的准确性,并调查对一系列广泛的非理想性因素的敏感性和稳健性。通过引入新的和高度现实性 AIMC的模型,我们大大改进了先前的再培训方法。我们发现,许多大型的大型的多层结构型DNNNNNNNN,包括非事实网络、经常的神经网络(RNMMS网络)和变式的内积积积质网络(RNIS网络)可以成功地显示我们的非结果。