![[2602.06142] Protean Compiler: An Agile Framework to Drive Fine-grain Phase Ordering](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.06142] Protean Compiler: An Agile Framework to Drive Fine-grain Phase Ordering
Summary
The Protean Compiler introduces an agile framework for fine-grain phase ordering in compilers, enhancing LLVM's capabilities with machine learning integration and achieving notable performance improvements.
Why It Matters
This research addresses a long-standing challenge in compiler optimization, providing a modern solution that leverages machine learning for better performance. By integrating phase ordering directly into LLVM, it offers a significant advancement for developers and researchers in programming languages and AI.
Key Takeaways
- Protean Compiler enhances LLVM with fine-grain phase ordering capabilities.
- Achieves performance speedups of up to 15.7% on specific applications.
- Integrates seamlessly with third-party ML frameworks and large language models.
- Offers over 140 handcrafted static feature collection methods.
- Promises open-source availability, fostering community collaboration.
Computer Science > Programming Languages arXiv:2602.06142 (cs) [Submitted on 5 Feb 2026 (v1), last revised 17 Feb 2026 (this version, v2)] Title:Protean Compiler: An Agile Framework to Drive Fine-grain Phase Ordering Authors:Amir H. Ashouri, Shayan Shirahmad Gale Bagi, Kavin Satheeskumar, Tejas Srikanth, Jonathan Zhao, Ibrahim Saidoun, Ziwen Wang, Bryan Chan, Tomasz S. Czajkowski View a PDF of the paper titled Protean Compiler: An Agile Framework to Drive Fine-grain Phase Ordering, by Amir H. Ashouri and 7 other authors View PDF HTML (experimental) Abstract:The phase ordering problem has been a long-standing challenge since the late 1970s, yet it remains an open problem due to having a vast optimization space and an unbounded nature, making it an open-ended problem without a finite solution, one can limit the scope by reducing the number and the length of optimizations. Traditionally, such locally optimized decisions are made by hand-coded algorithms tuned for a small number of benchmarks, often requiring significant effort to be retuned when the benchmark suite changes. In the past 20 years, Machine Learning has been employed to construct performance models to improve the selection and ordering of compiler optimizations, however, the approaches are not baked into the compiler seamlessly and never materialized to be leveraged at a fine-grained scope of code segments. This paper presents Protean Compiler: An agile framework to enable LLVM with built-in phase-ordering capabi...
