Secure Approximate Computing Systems

Approximate Computing (AC) techniques trade accuracy for performance improvement and energy efficiency, being increasingly attractive in various computation-intensive applications such as imaging processing, audio recognition, information search, and artificial intelligence. However, recent literature indicates that the utilization of AC techniques may raise new and unique security concerns. This project investigates hardware (HW) and software (SW) integrated methods to secure AC systems. 

Relevant publications:

  1. P. Yellu, and Q. Yu, “Can We Securely Use Approximate Computing?” to appear in Proc. IEEE International Symposium on Circuits and Systems (ISCAS 2020).
  2. P. Yellu, M. R. Monjur, T. Kammerer, D. Xu, and Q. Yu, “Security Threats and Countermeasures for Approximate Arithmetic Computing,” in Proc. Asia and South Pacific Design Automation Conference (ASP-DAC) 2020, pp. 259-264, Jan. 2020.
  3. P. Yellu, Z. Zhang, M. M. R. Monjur, R. Abeysinghe, and Q. Yu, “Emerging Applications of 3D Integration and Approximate Computing in High-Performance Computing Systems: Unique Security Vulnerabilities,” in Proc. 2019 IEEE High Performance Extreme Computing Conference, pp. 1-7, Sept. 2019.
  4. P. Yellu, N. Boskov , M. Kinsy, and Q. Yu, “Security Threats on Approximate Computing Systems,” in Proc. Great Lakes Symposium on VLSI (GLSVLSI'19), pp. 387-392, May 2019.
  5. L. Bu, J. Dofe, Q. Yu, and M. Kinsy, “SRASA: a Generalized Theoretical Framework for Security and Reliability Analysis in Computing Systems,” Journal of Hardware and Systems Security, pp. 1-19, Sept. 2018.