Hsinyu Tsai, Jed W. Pitera, et al.
ACS Nano
Recent advances in deep learning have been driven by ever-increasing model sizes, with networks growing to millions or even billions of parameters. Such enormous models call for fast and energy-efficient hardware accelerators. We study the potential of Analog AI accelerators based on Non-Volatile Memory, in particular Phase Change Memory (PCM), for software-equivalent accurate inference of natural language processing applications. We demonstrate a path to software-equivalent accuracy for the GLUE benchmark on BERT (Bidirectional Encoder Representations from Transformers), by combining noise-aware training to combat inherent PCM drift and noise sources, together with reduced-precision digital attention-block computation down to INT6.
Hsinyu Tsai, Jed W. Pitera, et al.
ACS Nano
Stefano Ambrogio, Pritish Narayanan, et al.
AICAS 2020
Y. Li, S. Kim, et al.
VLSI Technology 2018
Charles Mackin, Pritish Narayanan, et al.
CLEO/Europe-EQEC 2019