high performance computing on graphics processing units: hgpu.org

hgpu.org » Performance

KernelBlaster: Continual Cross-Task CUDA Optimization via Memory-Augmented In-Context Reinforcement Learning

Kris Shengjun Dong, Sahil Modi, Dima Nikiforov, Sana Damani, Edward Lin, Siva Kumar Sastry Hari, Christos Kozyrakis

View

Download (PDF)

Tags: Computer science, CUDA, LLM, nVidia, nVidia A100, nVidia H100, nVidia L40s, nVidia RTX A6000, Performance

February 23, 2026 by hgpu

OptiML: An End-to-End Framework for Program Synthesis and CUDA Kernel Optimization

Arijit Bhattacharjee, Heng Ping, Son Vu Le, Paul Bogdan, Nesreen K. Ahmed, Ali Jannesari

View

Download (PDF)

Tags: Computer science, CUDA, LLM, nVidia, nVidia A100, Performance

February 23, 2026 by hgpu

Execution-Centric Characterization of FP8 Matrix Cores, Asynchronous Execution, and Structured Sparsity on AMD MI300A

Aaron Jarmusch, Connor Vitz, Sunita Chandrasekaran

View

Download (PDF)

Tags: AMD, AMD Radeon Instinct MI300A, Benchmarking, Computer science, HIP, Performance, ROCm

February 16, 2026 by hgpu

Improving HPC Code Generation Capability of LLMs via Online Reinforcement Learning with Real-Machine Benchmark Rewards

Ryo Mikasa, Shun-ichiro Hayashi, Daichi Mukunoki, Tetsuya Hoshino, Takahiro Katagiri

View

Download (PDF)

Tags: Benchmarking, Code generation, Computer science, CUDA, HPC, LLM, Matrix multiplication, nVidia, nVidia H100, OpenMP, Performance

February 16, 2026 by hgpu

Deep Kernel Fusion for Transformers

Zixi Zhang, Zhiwen Mo, Yiren Zhao, Robert Mullins

View

Download (PDF)

Tags: Computer science, CUDA, LLM, nVidia, nVidia A100, Performance

February 16, 2026 by hgpu

Nsight Python: A Python-First Profiling Toolkit for Seamless GPU Kernel Analysis (Tool)

Bastian Hagedorn, Alexander Collins, Tony Mongkolsmai, Vinod Grover

View

Download (PDF)

Source codes

Tags: Computer science, CUDA, nVidia, nVidia B200, Package, Performance, Profiling, Python, Triton

February 2, 2026 by hgpu

ProfInfer: An eBPF-based Fine-Grained LLM Inference Profiler

Bohua Zou, Debayan Roy, Dhimankumar Yogesh Airao, Weihao Xu, Binqi Sun, Yutao Liu, Haibo Chen

View

Download (PDF)

Tags: Computer science, LLM, OpenCL, Performance

February 2, 2026 by hgpu

SynPerf: A Hybrid Analytical-ML Framework for GPU Performance Prediction

Kaixuan Zhang, Yunfan Cui, Shuhao Zhang, Chutong Ding, Shiyou Qian, Luping Wang, Jian Cao, Guangtao Xue, Cheng Huang, Guodong Yang, Liping Zhang

View

Download (PDF)

Tags: Computer science, CUDA, Heterogeneous systems, Machine learning, nVidia, nVidia A100, nVidia A40, nVidia H100, nVidia H20, nVidia H200, nVidia H800, nVidia L20, nVidia L40, nVidia RTX 6000 Ada, Performance, Triton

January 25, 2026 by hgpu

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

Ruifan Chu, Anbang Wang, Xiuxiu Bai, Shuai Liu, and Xiaoshe Dong

View

Download (PDF)

Tags: Computer science, CUDA, HIP, HPC, LLM, nVidia, nVidia GeForce RTX 2080 Ti, Performance

January 4, 2026 by hgpu

AccelOpt: A Self-Improving LLM Agentic System for AI Accelerator Kernel Optimization

Genghan Zhang, Shaowei Zhu, Anjiang Wei, Zhenyu Song, Allen Nie, Zhen Jia, Nandita Vijaykumar, Yida Wang, Kunle Olukotun

View

Download (PDF)

Tags: AI, Benchmarking, Computer science, LLM, Machine learning, Performance

December 29, 2025 by hgpu

tritonBLAS: Triton-based Analytical Approach for GEMM Kernel Parameter Selection

Ryan Swann, Muhammad Osama, Xiaohu Guo, Bryant Nelson, Lixun Zhang, Alex Brown, Yen Ong, Ali Yazdani, Sean Siddens, Ganesh Dasika, Alex Underwood

View

Download (PDF)

Source codes

Tags: AMD, AMD Radeon Instinct MI300X, AMD Radeon Instinct MI350X, ATI, BLAS, Computer science, HPC, Package, Performance, ROCm, Triton

December 7, 2025 by hgpu

The Anatomy of a Triton Attention Kernel

Burkhard Ringlein, Jan van Lunteren, Radu Stoica, Thomas Parnell

View

Download (PDF)

Tags: AMD Radeon Instinct MI250, AMD Radeon Instinct MI300X, ATI, Computer science, CUDA, DSL, HIP, LLM, nVidia, nVidia H100, Performance, Programming Languages, Triton

November 23, 2025 by hgpu

KernelGYM & Dr. Kernel: A distributed GPU environment and a collection of RL training methods to support RL for Kernel Generations

Dr. Kernel: Reinforcement Learning Done Right for Triton Kernel Generations

* * *

high performance computing on graphics processing units: hgpu.org

KernelBlaster: Continual Cross-Task CUDA Optimization via Memory-Augmented In-Context Reinforcement Learning

OptiML: An End-to-End Framework for Program Synthesis and CUDA Kernel Optimization

Execution-Centric Characterization of FP8 Matrix Cores, Asynchronous Execution, and Structured Sparsity on AMD MI300A

Improving HPC Code Generation Capability of LLMs via Online Reinforcement Learning with Real-Machine Benchmark Rewards

Deep Kernel Fusion for Transformers

Nsight Python: A Python-First Profiling Toolkit for Seamless GPU Kernel Analysis (Tool)

ProfInfer: An eBPF-based Fine-Grained LLM Inference Profiler

SynPerf: A Hybrid Analytical-ML Framework for GPU Performance Prediction

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

AccelOpt: A Self-Improving LLM Agentic System for AI Accelerator Kernel Optimization

tritonBLAS: Triton-based Analytical Approach for GEMM Kernel Parameter Selection

The Anatomy of a Triton Attention Kernel

Recent source codes

CL4SE: A Context Learning Benchmark For Software Engineering Tasks

CodeScaler: Scaling Code LLM Training and Test-Time Inference via Execution-Free Reward Models

A Safety Report on GPT-5.2, Gemini 3 Pro, Qwen3-VL, Grok 4.1 Fast, Nano Banana Pro, and Seedream 4.5

DICE: Diffusion Large Language Models Excel at Generating CUDA Kernels

KernelGYM & Dr. Kernel: A distributed GPU environment and a collection of RL training methods to support RL for Kernel Generations

Vortex-Optimized Light-weight Toolchain (VOLT)

SciDef: Automated Definition Extraction from Scientific Literature

bioagent-bench: Benchmark for evaluating LLM agents in bioinformatics

Benchmark suite for LLM inference on NVIDIA consumer GPUs

Theorizer: from the paper Generating Literature-Driven Scientific Discoveries at Scale

Most viewed papers (last 30 days)