5 Hits in 3.0 sec

MLGO: a Machine Learning Guided Compiler Optimizations Framework [article]

Mircea Trofin
2021 arXiv   pre-print
We propose MLGO, a framework for integrating ML techniques systematically in an industrial compiler -- LLVM.  ...  Leveraging machine-learning (ML) techniques for compiler optimizations has been widely studied and explored in academia.  ...  Background Machine Learning Techniques for Replacing Compiler Optimization Heuristics There are two characteristics that make reinforcement learning (RL) a suitable tool for replacing compiler optimization  ... 
arXiv:2101.04808v1 fatcat:jl7owbq5xvf5xmo3qhpksrqtrq

MLGOPerf: An ML Guided Inliner to Optimize Performance [article]

Amir H. Ashouri, Mostafa Elhoushi, Yuzhe Hua, Xiang Wang, Muhammad Asif Manzoor, Bryan Chan, Yaoqing Gao
2022 arXiv   pre-print
For the past 25 years, we have witnessed an extensive application of Machine Learning to the Compiler space; the selection and the phase-ordering problem.  ...  It employs a secondary ML model to generate rewards used for training a retargeted Reinforcement learning agent, previously used as the primary model by MLGO.  ...  [54] propose MLGO which was the first of a kind to provide an ML-guided optimization for a pass, i.e., inline optimization 2 .  ... 
arXiv:2207.08389v2 fatcat:hukrihw22beo5dvwbd22duoqz4

Generating GPU Compiler Heuristics using Reinforcement Learning [article]

Ian Colbert, Jake Daly, Norm Rubin
2021 arXiv   pre-print
We show that our machine learning-based compiler autotuning framework matches or surpasses the frame rates for 98% of graphics benchmarks with an average uplift of 1.6% up to 15.8%.  ...  In this paper, we developed a GPU compiler autotuning framework that uses off-policy deep reinforcement learning to generate heuristics that improve the frame rates of graphics applications.  ...  Acknowledgements We would like to thank Mike Bedy, Robert Gottlieb, Chris Reeve, Andrew Dupont, Karen Dintino, Peter Scannell and the rest of the AMD GPU compiler team for insightful discussions and infrastructure  ... 
arXiv:2111.12055v1 fatcat:2v3ekxzzfncr5cz6yuunslqspi

CompilerGym: Robust, Performant Compiler Optimization Environments for AI Research [article]

Chris Cummins, Bram Wasti, Jiadong Guo, Brandon Cui, Jason Ansel, Sahir Gomez, Somya Jain, Jia Liu, Olivier Teytaud, Benoit Steiner, Yuandong Tian, Hugh Leather
2021 arXiv   pre-print
We introduce CompilerGym, a set of environments for real world compiler optimization tasks, and a toolkit for exposing new optimization tasks to compiler researchers.  ...  Unlike in other domains, compiler and AI researchers do not have access to the datasets and frameworks that enable fast iteration and development of ideas, and getting started requires a significant engineering  ...  Other compiler research tools include OpenTuner [29] and YaCoS [49] , autotuning frameworks that include an ensemble of techniques for compiler optimizations; cTuning [50] , a framework for distributing  ... 
arXiv:2109.08267v2 fatcat:s2a3qrrk7zczflszoeta3ztl7q

Understanding and exploiting optimal function inlining

Theodoros Theodoridis, Tobias Grosser, Zhendong Su
2022 Proceedings of the 27th ACM International Conference on Architectural Support for Programming Languages and Operating Systems  
Inlining is a core transformation in optimizing compilers. It replaces a function call (call site) with the body of the called function (callee).  ...  We show a significant gap between the state-of-the-art strategy in LLVM and optimal inlining when optimizing for binary size, an important, deterministic metric independent of workload (in contrast to  ...  Machine learning for inlining: Machine learning has been suggested as an alternative to łhand-craftedž inlining heuristics.  ... 
doi:10.1145/3503222.3507744 fatcat:f3tsmwwr7fdhrlxnfwwpslrliy