COMPASS: A Unified Decision-Intelligence System for Navigating Performance Trade-off in HPC
Published in arXiv preprint arXiv:2604.22688, 2026
Recommended citation: Ankur Lahiry, Banooqa Banday, Yugesh Bhattarai, Tanzima Z Islam, and Mohammad Zaeed. (2026). "COMPASS: A Unified Decision-Intelligence System for Navigating Performance Trade-off in HPC." arXiv preprint arXiv:2604.22688. https://arxiv.org/abs/2604.22688
A unified decision-intelligence system for HPC environments that recommends performant configurations by balancing speed, cost, and reliability trade-offs with explainable outputs and uncertainty-aware ranking. Scales to traces with 1.3B samples (126 GB), achieving up to 100× faster training and 80× faster inference than state-of-the-art generative baselines.
Authors: Ankur Lahiry, Banooqa Banday, Yugesh Bhattarai, Tanzima Z Islam, and Mohammad Zaeed