AI Paper Atlas

Research map / Large language models / Agents, code and math

KernelFoundry: Hardware-Aware Evolutionary GPU Kernel Optimization

Nina Wiedemann, Quentin Leboutet, Michael Paulitsch, Diana Wofk, Benjamin Ummenhofer

ICML 2026 · Large language models › Agents, code and math · approach: LLMs for optimization and heuristics · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add gradient guidance to the evolutionary search (ablation)KernelFoundry (SYCL, 20 iters) · KernelBench L2 repr. set (SYCL)average speedup: 2.178 -> 2.539Improved
add meta-prompt evolution (ablation)KernelFoundry (SYCL, 20 iters) · KernelBench L2 repr. set (SYCL)geometric speedup: 1.600 -> 1.833Improved
add meta-prompt evolution (ablation)KernelFoundry (SYCL, 20 iters) · KernelBench L2 repr. set (SYCL)average speedup: 2.505 -> 2.539No clear effect
add template-based parameter optimization (2 iterations) (ablation)KernelFoundry (CUDA, o3-mini) · KernelBench L1 repr. setaverage speedup: 1.204 -> 1.241Improved
add template-based parameter optimization (ablation)KernelFoundry (CUDA, o3-mini) · KernelBench L2 repr. setaverage speedup: 2.051 -> 2.104Improved
add template-based parameter optimization (ablation)KernelFoundry (CUDA, GPT-o3/o4-mini/4.1) · Robust-kbench (12 tasks)geometric mean speedup: 5.295 -> 5.225No clear effect
use KernelFoundry instead of OpenEvolve at equal budgetSYCL kernel generation, 10 iters · KernelBench L2 repr. set (SYCL)average speedup: 1.483 -> 2.059Improved
use KernelFoundry instead of OpenEvolve with 40 iterationsSYCL kernel generation · KernelBench L2 repr. set (SYCL)average speedup: 2.535 -> 2.732Improved
optimize kernels on the target GPU rather than another GPUKernelFoundry kernels · KernelBench L2 repr. set, LNL targetaverage hardware-speedup: 1.537 (LNL-optimized on LNL)Improved
KernelFoundry instead of best published robust-kbench kernelsCUDA kernels · Robust-kbench (12 tasks)geometric mean speedup: 3.499 -> 5.295Improved
generated SYCL kernels vs oneDNN on matmul with relu post-opKernelFoundry · oneDNN comparisonspeedup: 0.35Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy