Modeling and Controlling Many-Core HPC Processors: an Alternative to PID and Moving Average Algorithms

TL;DR

This paper develops a detailed thermal and power model for HPC MPSoCs, analyzing coupling effects, and proposes a fuzzy control-based thermal capping strategy that outperforms PID controllers in managing temperature and power.

Contribution

It introduces a comprehensive thermal and power model considering real-world coupling effects and proposes a novel fuzzy control strategy to improve thermal management in HPC MPSoCs.

Findings

Up to 5x reduction in maximum temperature exceedance

Average 3.56% faster application runtime

Effective control of thermal and power coupling effects

Abstract

The race towards performance increase and computing power has led to chips with heterogeneous and complex designs, integrating an ever-growing number of cores on the same monolithic chip or chiplet silicon die. Higher integration density, compounded with the slowdown of technology-driven power reduction, implies that power and thermal management become increasingly relevant. Unfortunately, existing research lacks a detailed analysis and modeling of thermal, power, and electrical coupling effects and how they have to be jointly considered to perform dynamic control of complex and heterogeneous Multi-Processor System on Chips (MPSoCs). To close the gap, in this work, we first provide a detailed thermal and power model targeting a modern High Performance Computing (HPC) MPSoC. We consider real-world coupling effects such as actuators' non-idealities and the exponential relation between the dissipated power, the temperature state, and the voltage level in a single processing element. We analyze how these factors affect the control algorithm behavior and the type of challenges that they pose. Based on the analysis, we propose a thermal capping strategy inspired by Fuzzy control theory to replace the state-of-the-art PID controller, as well as a root-finding iterative method to optimally choose the shared voltage value among cores grouped in the same voltage domain. We evaluate the proposed controller with model-in-the-loop and hardware-in-the-loop co-simulations. We show an improvement over state-of-the-art methods of up to 5x the maximum exceeded temperature while providing an average of 3.56% faster application execution runtime across all the evaluation scenarios.