GraphPerf-RT: A Graph-Driven Performance Model for Hardware-Aware Scheduling of OpenMP Codes

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Pivezhandi, Mohammad, Banisharif, Mahdi, Bakhshan, Saeed, Saifullah, Abusayeed, Jannesari, Ali
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866911388545318912
author Pivezhandi, Mohammad
Banisharif, Mahdi
Bakhshan, Saeed
Saifullah, Abusayeed
Jannesari, Ali
author_facet Pivezhandi, Mohammad
Banisharif, Mahdi
Bakhshan, Saeed
Saifullah, Abusayeed
Jannesari, Ali
contents Autonomous AI agents on embedded platforms require real-time, risk-aware scheduling under resource and thermal constraints. Classical heuristics struggle with workload irregularity, tabular regressors discard structural information, and model-free reinforcement learning (RL) risks overheating. We introduce GraphPerf-RT, a graph neural network surrogate achieving deep learning accuracy at heuristic speeds (2-7ms). GraphPerf-RT is, to our knowledge, the first to unify task DAG topology, CFG-derived code semantics, and runtime context (per-core DVFS, thermal state, utilization) in a heterogeneous graph with typed edges encoding precedence, placement, and contention. Evidential regression with Normal-Inverse-Gamma priors provides calibrated uncertainty; we validate on makespan prediction for risk-aware scheduling. Experiments on three ARM platforms (Jetson TX2, Orin NX, RUBIK Pi) achieve R^2 = 0.81 on log-transformed makespan with Spearman rho = 0.95 and conservative uncertainty calibration (PICP = 99.9% at 95% confidence). Integration with four RL methods demonstrates that multi-agent model-based RL with GraphPerf-RT as the world model achieves 66% makespan reduction and 82% energy reduction versus model-free baselines, with zero thermal violations.
format Preprint
id arxiv_https___arxiv_org_abs_2512_12091
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle GraphPerf-RT: A Graph-Driven Performance Model for Hardware-Aware Scheduling of OpenMP Codes
Pivezhandi, Mohammad
Banisharif, Mahdi
Bakhshan, Saeed
Saifullah, Abusayeed
Jannesari, Ali
Machine Learning
C.4; D.1.3; I.2.6
Autonomous AI agents on embedded platforms require real-time, risk-aware scheduling under resource and thermal constraints. Classical heuristics struggle with workload irregularity, tabular regressors discard structural information, and model-free reinforcement learning (RL) risks overheating. We introduce GraphPerf-RT, a graph neural network surrogate achieving deep learning accuracy at heuristic speeds (2-7ms). GraphPerf-RT is, to our knowledge, the first to unify task DAG topology, CFG-derived code semantics, and runtime context (per-core DVFS, thermal state, utilization) in a heterogeneous graph with typed edges encoding precedence, placement, and contention. Evidential regression with Normal-Inverse-Gamma priors provides calibrated uncertainty; we validate on makespan prediction for risk-aware scheduling. Experiments on three ARM platforms (Jetson TX2, Orin NX, RUBIK Pi) achieve R^2 = 0.81 on log-transformed makespan with Spearman rho = 0.95 and conservative uncertainty calibration (PICP = 99.9% at 95% confidence). Integration with four RL methods demonstrates that multi-agent model-based RL with GraphPerf-RT as the world model achieves 66% makespan reduction and 82% energy reduction versus model-free baselines, with zero thermal violations.
title GraphPerf-RT: A Graph-Driven Performance Model for Hardware-Aware Scheduling of OpenMP Codes
topic Machine Learning
C.4; D.1.3; I.2.6
url https://arxiv.org/abs/2512.12091