high performance computing on graphics processing units: hgpu.org

hgpu.org » Performance

Safe, Seamless, And Scalable Integration Of Asynchronous GPU Streams In PETSc

Jacob Faibussowitsch, Mark F. Adams, Richard Tran Mills, Stefano Zampini, Junchao Zhang

View

Download (PDF)

Source codes

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

July 9, 2023 by hgpu

Optimization Techniques for GPU Programming

Pieter Hijma, Stijn Heldens, Alessio Sclocco, Ben van Werkhoven, Henri E. Bal

View

Download (PDF)

Tags: Computer science, CUDA, nVidia, OpenCL, Performance, survey

July 9, 2023 by hgpu

Improving Automatic Parallel Training via Balanced Memory Workload Optimization

Yujie Wang, Youhe Jiang, Xupeng Miao, Fangcheng Fu, Xiaonan Nie, Bin Cui

View

Download (PDF)

Tags: Computer science, Deep learning, nVidia, nVidia Titan RTX, Performance, PyTorch

July 9, 2023 by hgpu

Out-of-the-box library support for DBMS operations on GPUs

Harish Kumar Harihara Subramanian, Bala Gurumurthy, Gabriel Campero Durand, David Broneske, Gunter Saake

View

Download (PDF)

Source codes

Tags: Computer science, Databases, Hashing, Heterogeneous systems, nVidia, nVidia GeForce RTX 2080 Ti, nVidia V100, OpenCL, Package, Performance

July 2, 2023 by hgpu

Managing, Profiling, and Optimizing Heterogeneous GPU Workloads

James Gleeson

View

Download (PDF)

Source codes

Tags: Computer science, CUDA, Heterogeneous systems, Neural networks, nVidia, nVidia GeForce GTX 480, nVidia GeForce RTX 2080 Ti, nVidia Grid K1, Optimization, Package, Performance, Tesla V100, Thesis, Virtualization

July 2, 2023 by hgpu

Reducing branch divergence to speed up parallel execution of unit testing on GPUs

Taghreed Bagies, Wei Le, Jeremy Sheafer, Ali Jannesari

View

Download (PDF)

Source codes

Tags: Benchmarking, Computer science, CUDA, nVidia, Package, Performance, Tesla V100

June 18, 2023 by hgpu

GPUHarbor: Testing GPU Memory Consistency at Large

Reese Levine, Mingun Cho, Devon McKee, Andrew Quinn, Tyler Sorensen

View

Download (PDF)

Tags: AMD Radeon Pro 5500M, ATI, Computer science, Memory model, nVidia, nVidia GeForce RTX 3080, nVidia Tegra TX1, OpenCL, Performance

June 11, 2023 by hgpu

Optimization and Portability of a Fusion OpenACC-based FORTRAN HPC Code from NVIDIA to AMD GPUs

Igor Sfiligoi, Emily A. Belli, Jeff Candy, Reuben D. Budiardja

View

Download (PDF)

Tags: AMD Radeon Instinct MI250X, ATI, Computer science, Fortran, MPI, nVidia, nVidia A100, OpenACC, Performance

May 21, 2023 by hgpu

Towards Alignment of Parallelism in SYCL and ISO C++

Simon John Pennycook, Ben Ashbaugh, James Brodman, Michael Kinsner, Steffen Larsen, Greg Lueck, Roland Schulz, Michael Voss

View

Download (PDF)

Tags: Computer science, OpenCL, Performance, SYCL

May 14, 2023 by hgpu

Prediction of Performance and Power Consumption of GPGPU Applications

Gargi Alavani, Santonu Sarkar

View

Download (PDF)

Source codes

Tags: Computer science, CUDA, Energy-efficient computing, Java, Machine learning, nVidia, Package, Performance, PTX, Tesla K20, Thesis

May 14, 2023 by hgpu

Dynamically Finding Optimal Kernel Launch Parameters for CUDA Programs

Taabish Jeshani

View

Download (PDF)

Source codes

Tags: Benchmarking, Computer science, CUDA, nVidia, nVidia GeForce RTX 2070, Package, Performance, Thesis

May 7, 2023 by hgpu

Redwood: Flexible and Portable Heterogeneous Tree Traversal Workloads

Yanwen Xu, Ang Li, Tyler Sorensen

View

Download (PDF)

Source codes

Tags: Benchmarking, Computer science, CUDA, FPGA, Heterogeneous systems, HLS, Intel UHD 630, nVidia, nVidia Jetson AGX Xavier, nVidia Jetson Nano, Package, Performance, SYCL

May 7, 2023 by hgpu

MSCCL++: A GPU-driven communication stack for scalable AI applications

MSCCL++: Rethinking GPU Communication Abstractions for Cutting-edge AI Applications

Benchmark compute shader of Unity against InteropUnityCUDA

InteropUnityCUDA: A Tool for Interoperability Between Unity and CUDA

Data-efficient LLM Fine-tuning for Code Generation

DeepSpeed is a deep learning optimization library that makes distributed training and inference easy, efficient, and effective.

DeepCompile: A Compiler-Driven Approach to Optimizing Distributed Deep Learning Training

Large Language Model Powered C-to-CUDA Code Translation: A Novel Auto-Parallelization Framework

GigaAPI: a user-space API that simplifies multi-GPU programming, bridging the gap between the capabilities of parallel GPU systems and the ability of developers to harness their full potential

GigaAPI for GPU Parallelization

Coccinelle: a C code transformation engine using SmPL for matches, refactorings, and bug fixing

Advances in Semantic Patching for HPC-oriented Refactorings with Coccinelle

* * *

high performance computing on graphics processing units: hgpu.org

Safe, Seamless, And Scalable Integration Of Asynchronous GPU Streams In PETSc

Optimization Techniques for GPU Programming

Improving Automatic Parallel Training via Balanced Memory Workload Optimization

Out-of-the-box library support for DBMS operations on GPUs

Managing, Profiling, and Optimizing Heterogeneous GPU Workloads

Reducing branch divergence to speed up parallel execution of unit testing on GPUs

GPUHarbor: Testing GPU Memory Consistency at Large

Optimization and Portability of a Fusion OpenACC-based FORTRAN HPC Code from NVIDIA to AMD GPUs

Towards Alignment of Parallelism in SYCL and ISO C++

Prediction of Performance and Power Consumption of GPGPU Applications

Dynamically Finding Optimal Kernel Launch Parameters for CUDA Programs

Redwood: Flexible and Portable Heterogeneous Tree Traversal Workloads

Recent source codes

MSCCL++: A GPU-driven communication stack for scalable AI applications

Benchmark compute shader of Unity against InteropUnityCUDA

Data-efficient LLM Fine-tuning for Code Generation

DeepSpeed is a deep learning optimization library that makes distributed training and inference easy, efficient, and effective.

Large Language Model Powered C-to-CUDA Code Translation: A Novel Auto-Parallelization Framework

GigaAPI: a user-space API that simplifies multi-GPU programming, bridging the gap between the capabilities of parallel GPU systems and the ability of developers to harness their full potential

Coccinelle: a C code transformation engine using SmPL for matches, refactorings, and bug fixing

DuoReduce: MLIR's benchmark

Shamrock: Multi-GPU hydrodynamics for astrophysics

LLMPerf: GPU Performance Modeling meets Large Language Models

Most viewed papers (last 30 days)