PaFeDSE reduces processor design testing by nearly half
PaFeDSE: Pareto-Feedback-Guided Microarchitecture Design Space Exploration for Efficient Pareto Search
Hardware Architecture
Summary
Designing faster and more efficient computer processors means testing many possible designs, which can take a lot of time and computer power. The authors created PaFeDSE, a method that learns from previous tests to focus only on the most promising designs. This approach cuts the number of design tests needed by almost half and filters out repetitive or unhelpful options, speeding up the whole process. The method was tested on real processor workloads and showed big improvements in exploration speed and efficiency.
What this means in practice
- •For processor architects: Reduce time and computational resources needed to identify efficient processor designs that balance speed, power, and size.
- •For hardware design automation engineers: Improve automated tools for exploring hardware design options by focusing simulations on promising candidates, avoiding wasted effort.
Authors
Joonyong Park, Hyojeong Lee, Sejeong Yang, Joo-Young Kim
Abstract
Efficient microarchitecture design space exploration (DSE) is critical for optimizing performance, power, and area trade-offs in modern out-of-order processors. While machine-learning-based DSE significantly reduces the cost of exhaustively evaluating large design spaces, existing methods still waste expensive simulations on Pareto-miss candidates that do not improve the Pareto frontier and repeatedly process redundant candidates during iterative exploration. In this paper, we propose PaFeDSE, a Pareto-feedback-guided microarchitecture DSE framework for efficient Pareto search. PaFeDSE leverages Pareto-hit feedback to prioritize candidates likely to improve the frontier and Pareto-miss feedback to suppress exploration of unpromising local regions, while filtering redundant candidates that remain low priority across iterations. Evaluated on SPEC CPU 2017 workloads, PaFeDSE requires up to 49.0% fewer simulations than the strongest prior baseline to reach target hypervolume levels, while achieving a 7.1x reduction in exploration overhead through adaptive candidate filtering.