His paper, “Performance Benchmarking of Spin-Orbit Torque Magnetic RAM (SOT-MRAM) for Deep Neural Network (DNN) Accelerators,” presents a feasibility study on using a novel non-volatile memory technology. The technology, called spin-orbit torque magnetic RAM (SOT-MRAM) for AI accelerator, could potentially help build AI accelerators with better standby performance and a smaller chip area, compared to the design of current alternatives. These improvements are attributed to the lower leakage power and the projected smaller cell size.

The work presented was a collaboration with Luo’s Ph.D. advisor, associate professor Shimeng Yu, ECE Ph.D. candidates Piyush Kumar and Ching Liao (both advised by professor Azad Naeemi), as well as Standford professors Shan X. Wang (director of the Center for Magnetic Nanotechnology) and Wilman Tsai (executive director of Innovative Materials and Processes for Accelerated Computer Technologies research center).

This year’s IMW was held in Dresden, Germany from May 15-18, 2022. The IMW is the premier international forum for both new and seasoned technologists having diverse technical backgrounds to share and learn about the latest developments in memory technology with the global community.