A FeRAM based Volatile/Non-volatile Dual-mode Buffer Memory for Deep Neural Network Training
Yandong Luoa, Yuan-Chun Luo and Shimeng Yub
School of Electrical and Computer Engineering, Georgia Institute of Technology
ayandongluo@gatech.edu
bshimeng.yu@ece.gatech.edu
ABSTRACT
Deep neural network (DNN) training produces a large amount of intermediate data. As off-chip DRAM access is both energy and time consuming, sufficient on-chip buffer is preferred to achieve high energy efficiency for DNN accelerator designs. However, the low integration density and high leakage current of SRAM lead to large area cost and high standby power. The frequent refresh of embedded DRAM (eDRAM) degrades the energy efficiency due to its short refresh interval (40∼100μs). In this paper, a dual-mode buffer memory that can operate in both volatile eDRAM mode and non-volatile ferroelectric RAM (FeRAM) mode is proposed, which is based on the CMOS compatible HfZrO2 material. The functionality of the proposed dual-mode memory design is verified using SPICE simulation with the multi-domain Preisach model. A data lifetime-aware memory mode configuration protocol is proposed to optimize the buffer access energy. The architectural benchmark for DNN training shows 33.8%, 17.1% and 109.4% higher energy efficiency than baseline designs with eDRAM, FeRAM and SRAM with the same buffer area, respectively. The chip standby power is reduced by 26.8×∼47.5× and 1.5×∼10.6× compared with the SRAM and eDRAM baselines. The chip area overhead of the dual-mode buffer design is 5.7%.
Keywords: Deep Learning Accelerator, Embedded DRAM, Ferroelectric RAM, Refresh Optimization And Buffering.
