GPU Kernel Optimization Beyond Full Builds: An LLM Framework with Minimal Executable Programs

2025-12-15 · Ruifan Chu, Anbang Wang, Xiuxiu Bai, Shuai Liu, Xiaoshe Dong arxiv

In high-performance computing, hotspot GPU kernels are primary bottlenecks, and expert manual tuning is costly and hard to port. Large language model methods often assume kernels can be compiled and executed cheaply, which fails in large applications where full builds and runs are expensive. We present an end-to-end LLM framework with performance feedback that optimizes kernels without building the full application. From independently extracted hotspot kernels, it automatically completes code into a Minimal Executable Program (MEP), then performs multi-round iterative optimization and evaluation outside the full application. The framework integrates Automatic Error Repair and Performance Pattern Inheritance to fix faults, preserve correctness, reuse effective tiling/memory/synchronization strategies, and reduce search cost. Optimized variants are reintegrated into the original application for validation. We evaluate on NVIDIA GPUs and the Haiguang Deep Computing Unit (DCU) platform (AMD-licensed architecture) using PolyBench, the AMD APP SDK, and hotspot kernels from large-scale supercomputing applications. The method achieves average speedups of 5.05x (PolyBench on NVIDIA), 7.77x (PolyBench on DCU), 1.77x (AMD APP SDK), and 1.25x on three hotspot kernels, surpassing direct LLM optimization. The approach requires no full-source dependencies, offers cross-platform portability, and enables practical, low-cost GPU kernel optimization.

📄 PDF Abstract BibTeX arXiv:2512.22147

Code (0)

등록된 구현이 없습니다.

HTAM: Hierarchical Transition-Attended Memory for Operator Optimization

2026-05-28 · Yining Zhang, Mingyang Yi, Chen Wang, Xuwen Xiang 외 arxiv

High-performance GPU kernels are essential for efficient LLM deployment, yet optimizing them remains expertise-intensive. Recent LLM-based code generation makes automatic GPU operator generation promising, but operator o…

Code Generation

Lattice real-time simulations with learned optimal kernels

2023-10-12 · Daniel Alvestad, Alexander Rothkopf, Dénes Sexty

We present a simulation strategy for the real-time dynamics of quantum fields, inspired by reinforcement learning. It builds on the complex Langevin approach, which it amends with system specific prior information, a nec…

KernelFoundry: Hardware-aware evolutionary GPU kernel optimization

2026-03-12 · Nina Wiedemann, Quentin Leboutet, Michael Paulitsch, Diana Wofk 외 arxiv

Optimizing GPU kernels presents a significantly greater challenge for large language models (LLMs) than standard code generation tasks, as it requires understanding hardware architecture, parallel optimization strategies…

Code Generation

KernelEvolve: Scaling Agentic Kernel Coding for Heterogeneous AI Accelerators at Meta

2025-12-29 · Gang Liao, Hongsen Qin, Ying Wang, Alicia Golden 외 arxiv

Making deep learning recommendation model (DLRM) training and inference fast and efficient is important. However, this presents three key system challenges - model architecture diversity, kernel primitive diversity, and …

BOSS: Bayesian Optimization over String Spaces

2020-10-02 · NeurIPS 2020 12 · Henry B. Moss, Daniel Beck, Javier Gonzalez, David S. Leslie 외

This article develops a Bayesian optimization (BO) method which acts directly over raw strings, proposing the first uses of string kernels and genetic algorithms within BO loops. Recent applications of BO over strings ha…

Bayesian Optimization