High-speed multifunctional photonic memory on a foundry-processed photonic platform

作     者：Kari, Sadra rahimi Tamura, Marcus Guo, Zhimu Huang, Yi-siou Sun, Hongyi Lian, Chuanyu Nobile, Nicholas Erickson, John Moridsadat, Maryam Ocampo, C. A. R. L. O. S. A. RiOS Shastri, Bhavin J. Youngblood, Nathan 

作者机构：Univ Pittsburgh Dept Elect & Comp Engn Pittsburgh PA 15261 USA Queens Univ Ctr Nanophoton Dept Phys Engn Phys & Astron Kingston ON K7L 3N6 Canada Univ Maryland Dept Mat Sci & Engn College Pk MD 20742 USA Univ Maryland Inst Res Elect & Appl Phys College Pk MD 20742 USA 

出 版 物：《OPTICA》 (Optica)

年 卷 期：2025年第12卷第1期

页      面：31-38页

核心收录：

基　　金：Canada Foundation for Innovation Natural Sciences and Engineering Research Council of Canada Air Force Office of Scientific Research [FA9550-24-1-0064] National Science Foundation [2028624, 2210168, 2007774, 2329087, 2236972, 2210169] 

主　　题：Electron beam lithography Laser sources Neural networks Optical computing Optical confinement Quantum computation 

摘      要：The integration of computing with memory is essential for distributed, massively parallel, and adaptive architectures such as neural networks in artificial intelligence (AI). Accelerating AI can be achieved through photonic computing, but it requires nonvolatile photonic memory capable of rapid updates during on-chip training sessions or when new information becomes available during deployment. Phase-change materials (PCMs) are promising for providing compact, nonvolatile optical weighting;however, they face limitations in terms of bit precision, programming speed, and cycling endurance. Here, we propose a novel photonic memory cell that merges nonvolatile photonic weighting using PCMs with high-speed, volatile tuning enabled by an integrated PN junction. Our experiments demonstrate that the same PN modulator, fabricated via a foundry-compatible process, can achieve dual functionality. It supports coarse programmability for setting initial optical weights and facilitates high-speed fine-tuning to adjust these weights dynamically. The result shows a 400-fold increase in volatile tuning speed and a 10,000-fold enhancement in efficiency. This multifunctional photonic memory with volatile and nonvolatile capabilities could significantly advance the performance and versatility of photonic memory cells, providing robust solutions for dynamic computing environments. (c) 2025 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement