Automatic Kernel Generation with Large Language Models: Taxonomies, Evaluations, and Cross-Platform Challenges
Preprint
- 26 October 2026
- preprint
- Published by MDPI AG in Preprints.org
Abstract
Large language models (LLMs) have recently emerged as a promising tool for automatic generation and optimization of high-performance compute kernels. This rapidly growing area spans across hardware platforms and targets programming models such as AscendC, CUDA, and Triton. However, the literature remains fragmented, with substantial variation in generation workflows, training strategies, evaluation protocols, benchmark assumptions, hardware platforms, and reported metrics. This survey provides a structured review of LLM-based automatic kernel generation. We introduce a taxonomy based on the generation-time workflow, separating direct generation, refinement via structured feedback, search-based generation, and adaptive tool-orchestrating agentic systems. We further analyze orthogonal design axes including learning strategy, input representation, output representation, and hardware platform. Beyond method classification, we examine evaluation practices, including benchmark data, harnesses, integrity policies, as well as performance metrics, and discuss why reported results are often difficult to compare. Finally, we use Ascend NPU kernel generation as a cross-platform case study, highlighting challenges introduced by AscendC, host-device co-design, toolchain constraints, and limited public data. We conclude by identifying open problems in robust evaluation, higher-level task optimization, cross-platform transfer, production deployment, and benchmark design.Keywords
All Related Versions
This publication has 30 references indexed in Scilit:
- KernelEvolve: Scaling Agentic Kernel Coding for Heterogeneous AI Accelerators at Meta2026
- QiMeng-Kernel: Macro-Thinking Micro-Coding Paradigm for LLM-Based High-Performance GPU Kernel GenerationProceedings of the AAAI Conference on Artificial Intelligence, 2026
- QiMeng-GEMM: Automatically Generating High-Performance Matrix Multiplication Code by Exploiting Large Language ModelsProceedings of the AAAI Conference on Artificial Intelligence, 2025
- 2025 Reviewers—Consultants 2025Canadian Journal of Experimental Psychology / Revue canadienne de psychologie expérimentale, 2025
- TritonBench: Benchmarking Large Language Model Capabilities for Generating Triton Operators2025
- The Garbage Collection Handbook2023
- Triton: an intermediate language and compiler for tiled neural network computations2019
- Microsoft COCO: Common Objects in Context2014
- Dynamic storage allocation: A survey and critical review1995
- Optimization by Simulated AnnealingScience, 1983