Paper Title: Using Non-Volatile Memory for Computation-in-Memory

Paper Abstract: Non-volatile memory offers major opportunities for computation-in-memory (CIM). With the recent increasing interest in artificial intelligence (AI), academic researchers and startup companies are exploring the use of non-volatile memory technology (floating gate, MRAM, RRAM, and 3D Xpoint) to support CIM for AI applications. We explore the related design and manufacturing issues for various non-volatile technologies. When CIM performs machine learning, the non-volatile memory is subject to frequent write cycles. The analog circuits used for machine learning must be directly coupled with an array of non-volatile memory cells. Even minor design modifications could create design and implementation problems which will affect the testing, reliability, manufacturing, and final cost of products. Last, the major design issue is how to in-cooperate on chip Hardware Security?

Paper Author: WeiTi Liu, President, LucidPort Technology

Author Bio: Wei-Ti Liu is co-founder/President of LucidPort Technology, a maker of USB controllers and security devices. He was previously Co-Founder/GM/VP Engineering at PLX Technology, a maker of PCI-based chipsets. At PLX, he oversaw operations, directly managed the engineering group, and managed the foundry interface. He has extensive experience in ASIC VLSI chip designs and holds four US patents in that area. He has also been a design engineer for IBM, Intel, and AMD. He earned an MSEE from The City University of New York and a BSEE from the National Taiwan University. He has presented at several previous Flash Memory Summits.