BoGraph
- 5 April 2022
- conference paper
- conference paper
- Published by Association for Computing Machinery (ACM) in Proceedings of the 2nd European Workshop on Machine Learning and Systems
Abstract
Current auto-tuners struggle with computer systems due to their large complex parameter space and high evaluation cost. We propose BoGraph, an auto-tuning framework that builds a graph of the system components before optimizing it using causal structure learning. The graph contextualizes the system via decomposition of the parameter space for faster convergence and handling of many parameters. Furthermore, BoGraph exposes an API to encode experts' knowledge of the system via performance models and a known dependency structure of the components. We evaluated BoGraph via a hardware design case study achieving 5x -- 7x improvement in energy and latency over the default in a variety of tasks.Keywords
This publication has 17 references indexed in Scilit:
- Exploring Exploration: A Tutorial Introduction to Embedded Systems Design Space ExplorationIEEE Design & Test, 2016
- Co-designing accelerators and SoC interfaces using gem5-AladdinPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2016
- Taking the Human Out of the Loop: A Review of Bayesian OptimizationProceedings of the IEEE, 2015
- Aladdin: A pre-RTL, power-performance accelerator simulator enabling large design space exploration of customized architecturesPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2014
- Where do developers log? an empirical study on logging practices in industryPublished by Association for Computing Machinery (ACM) ,2014
- An evaluation of sequential model-based optimization for expensive blackbox functionsPublished by Association for Computing Machinery (ACM) ,2013
- Characterizing logging practices in open-source softwarePublished by Institute of Electrical and Electronics Engineers (IEEE) ,2012
- PetaBricksXRDS: Crossroads, The ACM Magazine for Students, 2010
- Energy-performance tradeoffs in processor architecture and circuit designACM SIGARCH Computer Architecture News, 2010
- Bayesian neural networks and density networksNuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, 1995