high performance computing on graphics processing units: hgpu.org

hgpu.org » Heterogeneous systems

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

Stuart H. Sul, Simran Arora, Benjamin F. Spector, Christopher Ré

View

Download (PDF)

Source codes

Tags: Computer science, CUDA, Heterogeneous systems, nVidia, nVidia B200, nVidia H100, Package

November 30, 2025 by hgpu

AIvailable: A Software-Defined Architecture for LLM-as-a-Service on Heterogeneous and Legacy GPUs

Pedro Antunes, Ana Rita Ortigoso, Gabriel Vieira, Daniel Fuentes, Luís Frazão, Nuno Costa, António Pereira

View

Download (PDF)

Tags: AMD Radeon Pro W6600, ATI, Computer science, CUDA, Heterogeneous systems, LLM, nVidia, nVidia GeForce GTX 1660, nVidia GeForce RTX 3070, ROCm

November 23, 2025 by hgpu

An MLIR pipeline for offloading Fortran to FPGAs via OpenMP

Gabriel Rodriguez-Canal, David Katz, Nick Brown

View

Download (PDF)

Source codes

Tags: Computer science, Fortran, FPGA, Heterogeneous systems, HLS, HPC, LLVM, OpenMP, Package

November 16, 2025 by hgpu

A High-Throughput GPU Framework for Adaptive Lossless Compression of Floating-Point Data

Zheng Li, Weiyan Wang, Ruiyuan Li, Chao Chen, Xianlei Long, Linjiang Zheng, Quanqing Xu, Chuanhui Yang

View

Download (PDF)

Source codes

Tags: Algorithms, Compression, Computer science, CUDA, Heterogeneous systems, nVidia, nVidia GeForce RTX 5080, Package

November 16, 2025 by hgpu

Enhancing Transformer Performance and Portability through Auto-tuning Frameworks

Patricia Siwinska,Jie Lei,Adrian Castello,Pedro Alonso-Jord́a,Enrique S. Quintana-Orti

View

Download (PDF)

Source codes

Tags: Auto-Tuning, Computer science, CUDA, Deep learning, Heterogeneous systems, LLM, nVidia, nVidia A100, Package, performance portability

November 2, 2025 by hgpu

Compiler and Runtime Systems for Generative AI Models

Zihao Ye

View

Download (PDF)

Source codes

Tags: Computer science, CUDA, Deep learning, Heterogeneous systems, Machine learning, nVidia, nVidia GeForce RTX 3070, nVidia H100, Package, Tesla V100

October 19, 2025 by hgpu

Accelerating cosmological simulations on GPUs: a portable approach using OpenMP

M. D. Lepinzan, G. Lacopo, D. Goz, G. Taffoni, P. Monaco, P. J. Elahi, U. Varetto, M. Cytowski

View

Download (PDF)

Source codes

Tags: AMD Radeon Instinct MI250X, Astrophysics, ATI, Benchmarking, Heterogeneous systems, HIP, HPC, Instrumentation and Methods for Astrophysics, nVidia, nVidia A100, OpenMP, Package

October 12, 2025 by hgpu

Scalable Engine and the Performance of Different LLM Models in a SLURM based HPC architecture

Anderson de Lima Luiz, Shubham Vijay Kurlekar, Munir Georges

View

Download (PDF)

Tags: Computer science, Heterogeneous systems, HPC, LLM, nVidia, nVidia A100

August 31, 2025 by hgpu

ConTraPh: Contrastive Learning for Parallelization and Performance Optimization

Quazi Ishtiaque Mahmud, Ali TehraniJamsaz, Nesreen K. Ahmed, Theodore L. Willke, Ali Jannesari

View

Download (PDF)

Tags: Code generation, Computer science, Heterogeneous systems, HPC, LLM, nVidia, OpenACC, OpenCL, OpenMP, Tesla V100

August 10, 2025 by hgpu

OpenDwarfs 2025: Modernizing the OpenDwarfs Benchmark Suite for Heterogeneous Computing

Juan José Ropero, Manuel de Castro, Diego R. Llanos

View

Download (PDF)

Source codes

Tags: AMD FirePro W7800, ATI, Benchmarking, Computer science, Heterogeneous systems, nVidia, nVidia RTX 4500 Ada, OpenCL, Package

August 3, 2025 by hgpu

Specx: a C++ task-based runtime system for heterogeneous distributed architectures

Paul Cardosi, Bérenger Bramas

View

Download (PDF)

Source codes

Tags: Computer science, CUDA, Heterogeneous systems, HIP, nVidia, nVidia A100, nVidia P100, nVidia Quadro RTX 8000, nVidia V100, Package, Task scheduling

July 20, 2025 by hgpu

Libra: Synergizing CUDA and Tensor Cores for High-Performance Sparse Matrix Multiplication

Jinliang Shi, Shigang Li, Youxuan Xu, Xueying Wang, Rongtian Fu, Zhi Ma, Tong Wu

View

Download (PDF)

Source codes

Tags: Computer science, CUDA, Deep learning, Heterogeneous systems, Matrix multiplication, nVidia, nVidia GeForce RTX 4090, nVidia H100, Package, Sparse matrix

July 6, 2025 by hgpu

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

ThunderKittens: Tile primitives for speedy kernels

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

NVIDIA Nemotron Parse 1.1

Iris: AMD RAD's multi-GPU Triton-based framework for seamless multi-GPU programming

Iris: First-Class Multi-GPU Programming Experience in Triton

HipKittens: Fast and Furious AMD Kernels

CudaForge: An Agent Framework with Hardware Feedback for CUDA Kernel Optimization

See all packages

* * *

high performance computing on graphics processing units: hgpu.org

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

AIvailable: A Software-Defined Architecture for LLM-as-a-Service on Heterogeneous and Legacy GPUs

An MLIR pipeline for offloading Fortran to FPGAs via OpenMP

A High-Throughput GPU Framework for Adaptive Lossless Compression of Floating-Point Data

Enhancing Transformer Performance and Portability through Auto-tuning Frameworks

Compiler and Runtime Systems for Generative AI Models

Accelerating cosmological simulations on GPUs: a portable approach using OpenMP

Scalable Engine and the Performance of Different LLM Models in a SLURM based HPC architecture

ConTraPh: Contrastive Learning for Parallelization and Performance Optimization

OpenDwarfs 2025: Modernizing the OpenDwarfs Benchmark Suite for Heterogeneous Computing

Specx: a C++ task-based runtime system for heterogeneous distributed architectures

Libra: Synergizing CUDA and Tensor Cores for High-Performance Sparse Matrix Multiplication

Recent source codes

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

hls4ml: Machine learning on FPGAs using HLS

ThunderKittens: Tile primitives for speedy kernels

NVIDIA Nemotron Parse 1.1

Iris: AMD RAD's multi-GPU Triton-based framework for seamless multi-GPU programming

HipKittens: Fast and Furious AMD Kernels

Fortran xDSL dialects

mt4g: Memory Topology 4 GPUs

Falcon: GPU-Based Floating-point Adaptive Lossless Compression

CudaForge: An Agent Framework with Hardware Feedback for CUDA Kernel Optimization

Most viewed papers (last 30 days)