PerfCoder: Large Language Models for Interpretable Code Performance Optimization

TL;DR

PerfCoder is a specialized large language model designed to generate interpretable, performance-optimized code by learning from real-world optimization data and reinforcement fine-tuning, significantly improving code speed and optimization strategies.

Contribution

Introducing PerfCoder, a fine-tuned LLM that produces interpretable, input-specific code optimizations, surpassing existing models in runtime speedup and effectiveness.

Findings

PerfCoder outperforms all existing models on the PIE benchmark.

It can generate interpretable feedback to guide further optimization.

Elevates 32B models and GPT-5 to new levels in code performance optimization.

Abstract

Large language models (LLMs) have achieved remarkable progress in automatic code generation, yet their ability to produce high-performance code remains limited--a critical requirement in real-world software systems. We argue that current LLMs struggle not only due to data scarcity but, more importantly, because they lack supervision that guides interpretable and effective performance improvements. In this work, we introduce PerfCoder, a family of LLMs specifically designed to generate performance-enhanced code from source code via interpretable, customized optimizations. PerfCoder is fine-tuned on a curated collection of real-world optimization trajectories with human-readable annotations, and preference-aligned by reinforcement fine-tuning using runtime measurements, enabling it to propose input-specific improvement strategies and apply them directly without relying on iterative refinement. On the PIE code performance benchmark, PerfCoder surpasses all existing models in both runtime speedup and effective optimization rate, demonstrating that performance optimization cannot be achieved by scale alone but requires optimization stratetgy awareness. In addition, PerfCoder can generate interpretable feedback about the source code, which, when provided as input to a larger LLM in a planner-and-optimizer cooperative workflow, can further improve outcomes. Specifically, we elevate the performance of 32B models and GPT-5 to new levels on code optimization, substantially surpassing their original performance.