Demonstration of transfer learning using 14 nm technology analog ReRAM array

• DOI: 10.3389/felec.2023.1331280
• Publication Date: 2024-01-15T04:32:28Z
• Authors: Fabia Farlin Athena; Omobayode Fagbohu...; Nanbo Gong; Malte J. Rasch; Jimmy Penaloza; SoonCheon Seo; Arthur Gasasira; Paul Solomon; Valeria Bragaglia; Steven Consiglio; Hisashi Higuchi; Chanro Park; Kevin Brew; Paul Jamison; Christopher Catano; Iqbal Saraf; Claire Silvestre; Xuefeng Liu; Babar Khan; Nikhil Jain; Steven McDermott; Rick Johnson; I. Estrada-Raygoza; Juntao Li; Tayfun Gokmen; Ning Li; Ruturaj Pujari; Fabio Carta; Hiroyuki Miyazoe; Martin M. Frank; Antonio La Porta; Devi Koty; Qingyun Yang; Robert D. Clark; Kandabara Tapily; Cory Wajda; Aelan Mosden; Jeff Shearer; Andrew Metz; Sean Teehan; Nicole Saulnier; Bert Offrein; Takaaki Tsunomura; Gert Leusink; Vijay Narayanan; Takashi Ando
• Keywords: Initialization; Transfer of learning; Edge device

Analog memory presents a promising solution in the face of growing demand for energy-efficient artificial intelligence (AI) at edge. In this study, we demonstrate efficient deep neural network transfer learning utilizing hardware and algorithm co-optimization an analog resistive random-access (ReRAM) array. For first time, illustrate that open-loop (DNN) image classification tasks, convergence rates can be accelerated by approximately 3.5 times through utilization co-optimized ReRAM hardware-aware Tiki-Taka v2 (TTv2) algorithm. A simulation based on statistical 14 nm CMOS array data provides insights into performance larger workloads, exhibiting notable improvement over conventional training with random initialization. This study shows DNN using optimized achieve faster smaller dataset compared to from scratch, thus augmenting AI capability