홈 › Papers

AsmEvo: Agentic Assembly-Level Optimization of AMD GPU Kernels with Functional Equivalence Verification

2026-08-21 · Ji Liu, Puyuan Yang, Rongzhang Zheng, Fan Wang, Jinglin Wang, Muhammad A. Awad, Mortis Huang, Andy Chang, Zekai Li, Zeping Li, Zihao An, Yue Liu, Yuchen Yang, Jianghui Wang, Chushi Chen, Ziqiong Liu, Fuwei Yang, Dong Li, Wen Heng Chung, Shengcai Liu, Emad Barsoum arxiv

High-performance ML systems increasingly rely on GPU kernels whose editable source is unavailable, generated, or too distant from final machine code to expose remaining optimizations. Existing LLM kernel optimizers and autotuners mainly operate on CUDA, Triton, HIP, or tensor-program source and validate against reference implementations. We study a stricter setting: optimizing an already compiled AMDGPU code object, where the deployed binary is the only behavioral oracle. We present AsmEvo, an agentic assembly-level optimizer for AMD GPU kernels. Given an AMDGPU code object K0, AsmEvo reconstructs a reassemblable representation, proposes low-level edits with a long-horizon agent, rebuilds an ABI-preserving optimized object, and accepts candidates only after differential verification against K0 under identical launches. AsmEvo combines code-object recovery, metadata-aware rebuilding, profiling-guided hot-window editing, correctness-gated timing, and conservative in-place patch fallback. We conduct extensive experiments with AsmEvo on various AMD GPU kernels. On MI308X, AsmEvo improves 29 of 30 selected KernelBench kernels, reaching 1.35x geometric-mean and 3.88x maximum speedup. On MI300X production workloads, it improves all evaluated AITer binaries and vLLM/SGLang Triton assembly kernels, reaching 1.09x/1.31x and 1.18x/1.34x geometric-mean/maximum speedups, respectively, while preserving functional equivalence.

📄 PDF Abstract BibTeX arXiv:2608.20711

Code (0)

등록된 구현이 없습니다.

ARGUS: Agentic GPU Optimization Guided by Data-Flow Invariants

2026-04-16 · Haohui Mai, Xiaoyan Guo, Xiangyun Ding, Daifeng Li 외 arxiv

LLM-based coding agents can generate functionally correct GPU kernels, yet their performance remains far below hand-optimized libraries on critical computations such as matrix multiplication, attention, and Mixture-of-Ex…

Reinforcement Learning

CuTeGen: An LLM-Based Agentic Framework for Generation and Optimization of High-Performance GPU Kernels using CuTe

2026-04-01 · Tara Saba, Zhiyang Chen, Jikai Jason Li, Anne Ouyang 외 arxiv

High-performance GPU kernels are critical to modern machine learning systems, yet developing them remains a manual, expert-driven process. Recent work has explored using LLMs to automate kernel generation, but generated …

CuAsmRL: Optimizing GPU SASS Schedules via Deep Reinforcement Learning

2025-01-14 · Guoliang He, Eiko Yoneki

Large language models (LLMs) are remarked by their substantial computational requirements. To mitigate the cost, researchers develop specialized CUDA kernels, which often fuse several tensor operations to maximize the ut…

Deep Reinforcement Learning GPU reinforcement-learning Reinforcement Learning+1

Towards Robust Agentic CUDA Kernel Benchmarking, Verification, and Optimization

2025-09-16 · Robert Tjarko Lange, Qi Sun, Aaditya Prasad, Maxence Faldor 외 arxiv

Recent advances in large language models (LLMs) demonstrate their effectiveness in scaling test-time compute for software engineering tasks. However, these approaches often focus on high-level solutions, with limited att…

Robotic Agentic Platform for Intelligent Electric Vehicle Disassembly

2026-03-19 · Zachary Allen, Max Conway, Lyle Antieau, Allen Ponraj 외 arxiv

Electric vehicles (EV) create an urgent need for scalable battery recycling, yet disassembly of EV battery packs remains largely manual due to high design variability. We present our Robotic Agentic Platform for Intellig…

Object Detection