The research, led by recent Ph.D. Graduate Xiaochen Peng and ECE Professor Shimeng Yu, proposes an end-to-end benchmark framework to evaluate state-of-the-art compute-in-memory (CIM) accelerators. The work provides the design automation community a family of open-sourced NeuroSim series with tools to allow users to explore different hardware specifications and find the best design options for their research on CIM accelerators.

The paper’s co-authors include Shanshi Huang, Hongwu Jiang, and Anni Lu, who are all current or former Ph.D. students in Yu’s Laboratory for Emerging Devices and Circuits research group.

The NeuroSim series has been recognized for its comprehensive capabilities and has attracted many users, including industry researchers from major companies such as Intel, TSMC, IBM, Samsung, and SK Hynix, as well as researchers from universities and academia worldwide. The project is primarily supported by the National Science Foundation and the Semiconductor Research Corporation.

The award will be presented at the Design Automation Conference (DAC), taking place from July 9-13 in San Francisco.

Earlier this year, Peng, who currently works at TSMC Corporate Research (North America) as a principal engineer, received the 2023 Outstanding Dissertation Award from the European Design and Automation Association (EDAA) for her NeuroSim series work.

In 2022, ECE Professor Sung Kyu Lim and his research team won the Donald O. Pederson Best Paper Award for their paper on a physical design tool named Compact-2D that automatically builds high-density and commercial-quality monolithic three-dimensional integrated circuits (3D ICs).