Protean Compiler integrates ML for fine-grained LLVM optimization

Researchers have developed Protean Compiler, an agile framework designed to integrate fine-grained phase ordering capabilities directly into the LLVM compiler. This framework includes over 140 static feature collection methods and aims to address the long-standing challenge of optimizing compiler phases. Experiments show speedup gains of up to 4.1% on average and up to 15.7% on select applications compared to LLVM's O3 optimization level, with minimal additional build time. The system also allows for integration with third-party ML frameworks and LLMs, demonstrating further speedup improvements on specific applications. AI

RANK_REASON This is a research paper detailing a new framework for compiler optimization. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• Large Language Models
• LLVM
• Machine Learning
• Protean Compiler

• paper
• infra

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →