Yingyan (Celine) Lin (Georgia Tech)
Title: LLM4AIGChip: Harnessing LLMs Towards the Automation of AI Accelerator Design
Abstract
In the rapidly evolving field of Artificial Intelligence (AI), the demand for efficient AI hardware accelerators is increasingly paramount. However, the complex and labor-intensive process of designing these accelerators presents significant challenges, hindering development. To address this, we have proposed an initiative called LLM4AIGChip, which aims to leverage the extraordinary capabilities of Large Language Models (LLMs) to revolutionize AI accelerator design and enhance accessibility. In this talk, I will share our progress on the LLM4AIGChip, demonstrating the potential of LLMs to facilitate the rapid development of AI accelerators, hardware design datasets, and design verification.