Variability-Aware Training and Self-Tuning of Highly Quantized DNNs for Analog PIM
Zihao Denga and Michael Orshanskyb
Department of Electrical and Computer Engineering University of Texas at Austin Austin TX, USA
azihaodeng@utexas.edu
borshansky@utexas.edu
ABSTRACT
DNNs deployed on analog processing in memory (PIM) architectures are subject to fabrication-time variability. We developed a new joint variability- and quantization-aware DNN training algorithm for highly quantized analog PIM-based models that is significantly more effective than prior work. It outperforms variability-oblivious and post-training quantized models on multiple computer vision datasets/models. For lowbitwidth models and high variation, the gain in accuracy is up to 35:7% for ResNet-18 over the best alternative.
We demonstrate that, under a realistic pattern of within- and between-chip components of variability, training alone is unable to prevent large DNN accuracy loss (of up to 54% on CIFAR- 100/ResNet-18). We introduce a self-tuning DNN architecture that dynamically adjusts layer-wise activations during inference and is effective in reducing accuracy loss to below 10%.
