LLM for SoC Security: A Paradigm Shift

Authors

Dipayan Saha, Shams Tarek, Katayoon Yahyaei, Sujan Kumar Saha, Jingbo Zhou, Mark Tehranipoor, Farimah Farahmandi

Abstract

Ensuring the security of system-on-chips (SoCs) and edge devices is critical, but traditional verification methods struggle with automation, scalability, comprehensiveness, and adaptability. This paper explores leveraging the capabilities of Large Language Models (LLMs), particularly generative pre-trained transformers (GPTs), to bridge these gaps and establish a more efficient, scalable, and adaptable methodology for SoC security verification. By integrating LLMs into the SoC security verification paradigm, we aim to open new possibilities and address the complexities of securing advanced SoCs. The study explores four key security tasks: vulnerability insertion, security assessment, security verification, and countermeasure development.

@article{saha2024llm,
  title={LLM for SoC Security: A Paradigm Shift},
  author={Saha, Dipayan and Tarek, Shams and Yahyaei, Katayoon and Saha, Sujan Kumar and Zhou, Jingbo and Tehranipoor, Mark and Farahmandi, Farimah},
  journal={IEEE Access},
  volume={12},
  pages={155498--155521},
  year={2024},
  publisher={IEEE},
  doi={10.1109/ACCESS.2024.3427369}
}