DataKernelBench: Can LLMs Optimize Database Queries on GPUs?

arXiv:2608.25061 · cs.CL, cs.AI, cs.DB, cs.LG, cs.PL · Submitted 2026-08-25 · Read on arXiv

cs.CL, cs.AI, cs.DB, cs.LG, cs.PL

Submitted: 2026-08-25

Updated: 2026-08-27

Code: https://github.com/kerneldf/datakernelbench

Project page: https://kerneldf.github.io/datakernelbench

Terminology

• Benchmark — 31× in this paper · explained in 2 episodes
• Core — 31× in this paper · explained in 2 episodes
• Aggregation — 7× in this paper · explained in 1 episode
• Ablation — 4× in this paper · explained in 1 episode
• Benchmarking — 3× in this paper · explained in 4 episodes
• Text-to-SQL — 3× in this paper · explained in 3 episodes
• International Data Corporation — 2× in this paper
• Tensor Query Processing — 2× in this paper
• Abstraction — 1× in this paper · explained in 1 episode
• Adam — 1× in this paper · explained in 2 episodes
• Agentic Workflows — 1× in this paper · explained in 1 episode
• Failure Modes — 1× in this paper · explained in 3 episodes
• Foundation Model — 1× in this paper · explained in 3 episodes
• Foundation Models — 1× in this paper · explained in 7 episodes
• Generalization — 1× in this paper · explained in 13 episodes
• Robustness — 1× in this paper · explained in 5 episodes
• Verification — 1× in this paper · explained in 2 episodes

Sources

• Constitutional AI: Harmlessness from AI Feedback
• CUDA Agent: Large-Scale Agentic RL for High-Performance CUDA Kernel Generation
• LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Code
• Towards Robust Agentic CUDA Kernel Benchmarking, Verification, and Optimization
• KernelEvolve: Scaling Agentic Kernel Coding for Heterogeneous AI Accelerators at Meta
• Evaluating Language Models for Efficient Code Generation
• PystachIO: Efficient Distributed GPU Query Processing with PyTorch over Fast Networks & Fast Storage
• Learned Query Superoptimization
• GPT-4 Technical Report
• gpt-oss-120b & gpt-oss-20b Model Card
• Devstral: Fine-tuning Language Models for Coding Agent Applications
• Bespoke OLAP: Synthesizing Workload-Specific One-size-fits-one Database Engines
• MultiKernelBench: A Multi-Platform Benchmark for Kernel Generation
• TritonRL: Training LLMs to Think and Code Triton Without Cheating
• GLM-4.5: Agentic, Reasoning, and Coding (ARC) Foundation Models

Related papers

• Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
• Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
• Subliminal Steering: Stronger Encoding of Hidden Signals
• MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
• The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
• Untangling the Mechanisms of Misleading Context in Medical Question Answering