Generating GPU Compiler Heuristics using Reinforcement Learning

2021/11/23 by Ian Colbert, Colbert, Ian, Jake Daly +3 · 1 citation
Computer Science · #Artificial Intelligence (cs.AI) #Computation and Language (cs.CL) #FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning and Data Classification #Reinforcement Learning in Robotics #Software Engineering Research

paper · pdf · doi:10.48550/arxiv.2111.12055

openalex publication_date 2021/11/23 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

GPU compilers are complex software programs with many optimizations specific to target hardware. These optimizations are often controlled by heuristics hand-designed by compiler experts using time- and resource-intensive processes. 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. Furthermore, we demonstrate the resilience of these learned heuristics to frequent compiler updates by analyzing their stability across a year of code check-ins without retraining. 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%.

Citations

• Device Placement Optimization with Reinforcement Learning

Cited by

• Discovering Software Parallelization Points Using Deep Neural Networks

Related

• Theoretical Foundations of GPU-Native Compilation for Rapid Code Iteration
• CuAsmRL: Optimizing GPU SASS Schedules via Deep Reinforcement Learning
• Autotuning GPU Kernels via Static and Predictive Analysis
• SIP: Autotuning GPU Native Schedules via Stochastic Instruction Perturbation
• Fine-Tuning GPT-5 for GPU Kernel Generation