Self-Adaptive Memory Approximation: A Formal Control Theory Approach

• Published in: IEEE embedded systems letters Vol. 12; no. 2; pp. 33 - 36
• Main Authors: Maity, Biswadip, Shoushtari, Majid, Rahmani, Amir M., Dutt, Nikil
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 01.06.2020; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Adaptive control; Approximate computing; Approximation; Bit error rate; Computer architecture; Control theory; Image edge detection; Knobs; Manuals; Mathematical analysis; Mathematical model; memory hierarchy; Quality; Quality control; Quality of service; Settling; System identification; Tuning
• ISSN: 1943-0663; 1943-0671
• DOI: 10.1109/LES.2019.2941018

Memory approximation enables trading off quality/accuracy for performance or energy gains. Traditionally, application programmers are burdened with the difficult task of setting memory approximation knobs to achieve the desired quality of service (QoS). Our self-adaptive approach for memory approximation eases the programmer's burden: simply specify the desired quality as a goal, with the system deploying a formal control-theoretic approach to tune the memory approximation knobs and deliver a guaranteed QoS. We model quality configuration tracking as a formal quality control problem, and outline a system identification technique that captures memory approximation effects with variations in application input and system architecture. Preliminary results show that we can alleviate the programmer's burden of manual knob tuning for on-chip memory approximation. When compared with a manual calibration scheme we achieve <inline-formula> <tex-math notation="LaTeX">3\times </tex-math></inline-formula> improvement in average settling time and up to <inline-formula> <tex-math notation="LaTeX">5\times </tex-math></inline-formula> improvement in best case settling time.