EECS Professors Receive a $200,000 NSF Award to Develop an AI-Driven Computer Architecture Learning Tool

The CHIPS Act in the US aims to train skilled computer architects for innovative computer hardware design. Computer system design relies heavily on software-based simulation, also used for teaching computer architecture concepts. However, current simulators used for education have steep learning curves, making them less accessible to beginners and prone to errors. 

EECS assistant professors Mohammad Alian and Tamzidul Hoque received a $200,000 award from the NSF to develop Scaffolded AI-driven Learning Simulation (SAILS), a novel framework and technology that creates an interactive learning platform for computer architecture and offers design exercises at different difficulty levels. It aims to make learning computer architecture more engaging and accessible to a wider audience.

SAILS uses AI-driven technology to reduce the complexity of computer architecture simulators for educational purposes. Its front-end simplifies simulation for students with varying levels of background knowledge. The back-end connects to a state-of-the-art architecture simulator and provides personalized assistance to users through a centralized AI model. The model learns from individual and team data, as well as the collective experience of all users. SAILS also employs the "faded scaffolding" approach to provide appropriate levels of support to maximize learning for individual learners and teams.