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.

A Power Modelling Approach for Many-Core Architectures.

Power Profiling for Low Power Optimization Using New Computing Model

Multi-core Chip Dynamic Power Management Framework Based on Reinforcement Learning br

Towards Modeling Energy Consumption of Xeon Phi

PoweRock: Power Modeling and Flexible Dynamic Power Management for Many-Core Architectures.

PMCC: Fast and Accurate System-Level Power Modeling for Processors on Heterogeneous SoC

Cross Architectural Power Modelling

Performance and power modeling and prediction using MuMMI and 10 machine learning methods

A Fast Approach For System-Level Power Modeling And Simulation

The Power-Performance Tradeoffs of the Intel Xeon Phi on HPC Applications

HaPPy: Hyperthread-Aware Power Profiling Dynamically

Power-Efficient Schemes Via Workload Characterization On The Intel'S Single-Chip Cloud Computer

Characterizing Power Delivery Systems with On/off-Chip Voltage Regulators for Many-Core Processors

A Silicon-Validated Methodology for Power Delivery Modeling and Simulation

Efficient and Precise Profiling, Modeling and Management on Power and Performance for Power Constrained HPC Systems

Performance and Power Modeling and Prediction Using MuMMI and Ten Machine Learning Methods

Prediction models for multi-dimensional power-performance optimization on many cores

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

Accurate CPU Power Modeling for Multicore Smartphones

Dynamic Power Estimation with Hardware Performance Counters Support on Multi-core Platform

Multi-node Power/Performance Modeling for HPC System