Special Session Ⅱ

Compute-in-Memory for Embodied AI

Chair		Co-Chair
An Guo City University of Hong Kong, China		Jintao Li University of Electronic Science and Technology of China, China

Keywords: Compute-in-Memory (CIM), AI accelerator, Embodied AI, Neural Processing Unit (NPU), CIM Electronic Design Automation (EDA)

Special Session Information: 

This session targets the perception-decision-control loop of embodied intelligence, focusing on key challenges and recent advances of Compute-in-Memory (CIM) for edge agents. 

Topics of interest include but are not limited to:

• CIM Macro

• CIM-based accelerator

• CIM-based algorithm-circuit-device co-design

• FP/INT quantization for CIM

• CIM system operator mapping

• CIM system prototypes and benchmarking

• CIM architecture design space exploration