high performance computing on graphics processing units: hgpu.org

Yinan Li, Bailu Ding, Ziyun Wei, Lukas Maas, Momin Al-Ghosien, Spyros Blanas, Nicolas Bruno, Carlo Curino, Matteo Interlandi, Craig Peeper, Kaushik Rajan, Surajit Chaudhuri, Johannes Gehrke

View

Download (PDF)

Tags: Benchmarking, Computer science, Databases, Memory, nVidia, nVidia A100, nVidia H100

August 31, 2025 by hgpu

Performant Unified GPU Kernels for Portable Singular Value Computation Across Hardware and Precision

Evelyne Ringoot, Rabab Alomairy, Valentin Churavy, Alan Edelman

View

Download (PDF)

Tags: AMD Radeon Instinct MI250, Apple M1 Pro, ATI, Computer science, HIP, Intel, Intel Ponte Vecchio Max 1100, Kokkos, Linear Algebra, Machine learning, nVidia, nVidia A100, nVidia GeForce RTX 4060, nVidia H100, OpenCL, SYCL

August 17, 2025 by hgpu

Performance Portable Gradient Computations Using Source Transformation

Kim Liegeois, Brian Kelley, Eric Phipps, Sivasankaran Rajamanickam, Vassil Vassilev

View

Download (PDF)

Tags: AMD Radeon Instinct MI300X, ATI, Computer science, CUDA, HIP, Kokkos, Machine learning, Mathematical Software, nVidia, nVidia H100

August 3, 2025 by hgpu

Kevin: Multi-Turn RL for Generating CUDA Kernels

Carlo Baronio, Pietro Marsella, Ben Pan, Simon Guo, Silas Alberti

View

Download (PDF)

Tags: Artificial intelligence, Computer science, CUDA, LLM, Machine learning, nVidia, nVidia H100, nVidia H200, Performance

July 20, 2025 by hgpu

CUDA-L2: Surpassing cuBLAS Performance for Matrix Multiplication through Reinforcement Learning

BoltzGen:Toward Universal Binder Design

cuPilot: A Strategy-Coordinated Multi-agent Framework for CUDA Kernel Evolution

MATLAB Tensor Core models

Accurate Models of NVIDIA Tensor Cores

TritonForge: Transform PyTorch Operations into Optimized GPU Kernels with LLMs

TritonForge: Profiling-Guided Framework for Automated Triton Kernel Optimization

RLTune: Hybrid Learning and Optimization-Based Dynamic Scheduling for DL Workloads on Heterogeneous GPU Clusters

Hybrid Learning and Optimization-Based Dynamic Scheduling for DL Workloads on Heterogeneous GPU Clusters

tritonBLAS: A Lightweight Triton-based General Matrix Multiplication (GEMM) Library

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

hls4ml: Machine learning on FPGAs using HLS

hls4ml: A Flexible, Open-Source Platform for Deep Learning Acceleration on Reconfigurable Hardware

NVIDIA Nemotron Parse 1.1

ThunderKittens: Tile primitives for speedy kernels

ParallelKittens: Systematic and Practical Simplification of Multi-GPU AI Kernels

See all packages

* * *

high performance computing on graphics processing units: hgpu.org

Tutoring LLM into a Better CUDA Optimizer

Compiler and Runtime Systems for Generative AI Models

A Performance Portable Matrix Free Dense MTTKRP in GenTen

EvoEngineer: Mastering Automated CUDA Kernel Code Evolution with Large Language Models

Opal: A Modular Framework for Optimizing Performance using Analytics and LLMs

Mojo: MLIR-Based Performance-Portable HPC Science Kernels on GPUs for the Python Ecosystem

Towards Robust Agentic CUDA Kernel Benchmarking, Verification, and Optimization

Combining Performance and Productivity: Accelerating the Network Sensing Graph Challenge with GPUs and Commodity Data Science Software

Scaling GPU-Accelerated Databases beyond GPU Memory Size

Performant Unified GPU Kernels for Portable Singular Value Computation Across Hardware and Precision

Performance Portable Gradient Computations Using Source Transformation

Kevin: Multi-Turn RL for Generating CUDA Kernels

Recent source codes

CUDA-L2: Surpassing cuBLAS Performance for Matrix Multiplication through Reinforcement Learning

BoltzGen:Toward Universal Binder Design

cuPilot: A Strategy-Coordinated Multi-agent Framework for CUDA Kernel Evolution

MATLAB Tensor Core models

TritonForge: Transform PyTorch Operations into Optimized GPU Kernels with LLMs

RLTune: Hybrid Learning and Optimization-Based Dynamic Scheduling for DL Workloads on Heterogeneous GPU Clusters

tritonBLAS: A Lightweight Triton-based General Matrix Multiplication (GEMM) Library

hls4ml: Machine learning on FPGAs using HLS

NVIDIA Nemotron Parse 1.1

ThunderKittens: Tile primitives for speedy kernels

Most viewed papers (last 30 days)