Storing Analog Data with Analog-Valued Non-Volatile Memory Array: An Adaptive Error-Resilient Joint Source-Channel Co
Presenter: Xin Zheng (Stanford)
Abstract: Data stored in the cloud or on mobile devices reside in physical memory systems with finite sizes. Today, huge amounts of analog data, e.g. images, sounds and videos, are first converted into binary digital form and then information compression algorithms (source coding) and error-correcting codes (channel coding) are separately employed to minimize the amount of physical storage required. Emerging non-volatile memory (NVM) technologies (e.g., Phase-change Memory (PCM) and resistive RAM (RRAM)) provide the possibility to store the analog information directly into analog memory system. In this talk, I will present an adaptive joint source channel coding scheme developed with a neural network for efficient storage of analog data onto NVM arrays. We demonstrate with experiments an image compression and storage task with analog-valued PCM array and RRAM array. With a cross-layer design from algorithm to device technology, the scheme is shown to exploit the full analog dynamic range of PCM and RRAM devices and be resilient to the device non-idealities (e.g. defective cells, device variability, resistance drift and relaxation) at the same time.
Bio: Xin Zheng is a PhD candidate in Electrical Engineering at Stanford University, supervised by Prof. H. -S. Philip Wong. She received her B.S. degree in Physics from Nanjing University, and her M.S. degree in Electrical Engineering from Stanford University. Her PhD research focuses on analog storage system enabled by emerging memory technologies (e.g., RRAM and PCM).
Theme 4, Task 2776.072: Materials and Devices for Achieving Analog Updates for Online Training and Inferencing