high performance computing on graphics processing units: hgpu.org

Applications

hgpu.org » Applications » Computer science

Toward Performance Portability for CPUs and GPUs Through Algorithmic Compositions

Li-Wen Chang

View

Tags: Algorithms, Computer science, Heterogeneous systems, nVidia, OpenCL, performance portability, Tesla C2050, Tesla K20, Thesis

October 15, 2017 by hgpu

SoAx: A generic C++ Structure of Arrays for handling Particles in HPC Codes

Holger Homann, Francois Laenen

View

Download (PDF)

Source codes

Tags: Benchmarking, Computational Physics, Computer science, CUDA, Heterogeneous systems, nVidia, nVidia GeForce GT 755 M, Package, Performance, Physics, Tesla M2050

October 15, 2017 by hgpu

Synkhronos: a Multi-GPU Theano Extension for Data Parallelism

Adam Stooke, Pieter Abbeel

View

Download (PDF)

Source codes

Tags: Computer science, CUDA, Data parallelism, Deep learning, nVidia, nVidia DGX-1, Package, Tesla P100, Theano

October 15, 2017 by hgpu

Energy efficiency of finite difference algorithms on multicore CPUs, GPUs, and Intel Xeon Phi processors

Satya P. Jammy, Christian T. Jacobs, David J. Lusher, Neil D. Sandham

View

Download (PDF)

Tags: Algorithms, Computer science, CUDA, Energy-efficient computing, Finite difference, Intel Xeon Phi, nVidia, Tesla K40

October 3, 2017 by hgpu

FPGA implementation of a Convolutional Neural Network for "Wake up word" detection

Ole Martin Skafsa

View

Download (PDF)

Tags: CNN, Computer science, Deep learning, FPGA, Machine learning, Neural networks, NLP, nVidia, nVidia Tegra TX1, OpenCL, Thesis

October 3, 2017 by hgpu

An Efficient Load Balancing Method for Tree Algorithms

Osama Talaat Ibrahim, Ahmed El-Mahdy

View

Download (PDF)

Tags: Algorithms, Computer science, Intel Xeon Phi, load balancing, Performance

October 3, 2017 by hgpu

Computing Treewidth on the GPU

Tom C. van der Zanden, Hans L. Bodlaender

View

Download (PDF)

Source codes

Tags: Algorithms, Bloom filter, Computer science, Graph theory, nVidia, nVidia GeForce GTX 1060, OpenCL, Package

October 3, 2017 by hgpu

Performance Evaluation of Container-based Virtualization for High Performance Computing Environments

Carlos Arango, Remy Dernat, John Sanabria

View

Download (PDF)

Source codes

Tags: Benchmarking, Computer science, CUDA, MPI, nVidia, Operating systems, Package, Performance, Tesla K20, Virtualization

October 3, 2017 by hgpu

OpenCL Actors – Adding Data Parallelism to Actor-based Programming with CAF

Raphael Hiesgen, Dominik Charousset, Thomas C. Schmidt

View

Download (PDF)

Source codes

Tags: Computer science, Data parallelism, Heterogeneous systems, Intel Xeon Phi, nVidia, nVidia GeForce GTX 780 M, OpenCL, Package, Tesla C2075

September 28, 2017 by hgpu

Modeling the Resource Requirements of Convolutional Neural Networks on Mobile Devices

Zongqing Lu, Swati Rallapalli, Kevin Chan, Thomas La Porta

View

Download (PDF)

Tags: Computer science, Computer vision, CUDA, Deep learning, Neural networks, nVidia, nVidia Jetson TK1, nVidia Tegra TX1

September 28, 2017 by hgpu

Accelerating Electron Tomography Reconstruction Algorithm ICON Using the Intel Xeon Phi Coprocessor on Tianhe-2 Supercomputer

Zihao Wang, Yu Chen, Jingrong Zhang, Lun Li, Xiaohua Wan, Zhiyong Liu, Fei Sun, Fa Zhang

View

Download (PDF)

Tags: intel mic, Intel Xeon Phi, NUFFT, supercomputer, Tianhe-2

September 28, 2017 by holy

Asynchronous Task-Based Polar Decomposition on Single Node Manycore Architectures

Dalal Sukkari, Hatem Ltaief, Mathieu Faverge, David Keyes

View

Download (PDF)

Tags: Algorithms, Benchmarking, Computer science, Factorization, Intel Xeon Phi, nVidia, StarPU, Task scheduling, Tesla K80, Tesla P100

September 21, 2017 by hgpu

CUDAnalyst (CUDA + Analyst)

Towards Feedback-to-Plan Decisions for Self-Evolving LLM Agents in CUDA Kernel Generation

CodegenBench

CodegenBench: Can LLMs Write Efficient Code Across Architectures?

KernelBenchX: A Comprehensive Benchmark for Evaluating LLM-Generated GPU Kernels

CuTile Benchmark Suite: Performance and Productivity Tradeoffs for GPU Kernel Programming on Blackwell Architecture

Evaluating CUDA Tile for AI Workloads on Hopper and Blackwell GPUs

Device Virtual Machine (DVM)

DVM: Real-Time Kernel Generation for Dynamic AI Models

MegaTrain: Full Precision Training of 100B+ Parameter Large Language Models on a Single GPU

Agentic Code Optimization via Compiler-LLM Cooperation

See all packages

* * *

high performance computing on graphics processing units: hgpu.org

Applications

Toward Performance Portability for CPUs and GPUs Through Algorithmic Compositions

SoAx: A generic C++ Structure of Arrays for handling Particles in HPC Codes

Synkhronos: a Multi-GPU Theano Extension for Data Parallelism

Energy efficiency of finite difference algorithms on multicore CPUs, GPUs, and Intel Xeon Phi processors

FPGA implementation of a Convolutional Neural Network for "Wake up word" detection

An Efficient Load Balancing Method for Tree Algorithms

Computing Treewidth on the GPU

Performance Evaluation of Container-based Virtualization for High Performance Computing Environments

OpenCL Actors – Adding Data Parallelism to Actor-based Programming with CAF

Modeling the Resource Requirements of Convolutional Neural Networks on Mobile Devices

Accelerating Electron Tomography Reconstruction Algorithm ICON Using the Intel Xeon Phi Coprocessor on Tianhe-2 Supercomputer

Asynchronous Task-Based Polar Decomposition on Single Node Manycore Architectures

Recent source codes

CUDAnalyst (CUDA + Analyst)

CodegenBench

KernelBenchX: A Comprehensive Benchmark for Evaluating LLM-Generated GPU Kernels

CUDA Kernel Fusion Benchmarks

IntelliKit: Agent-first tooling for AMD hardware

DITRON: Distributed Compiler based on Triton for Parallel Systems

CuTile Benchmark Suite: Performance and Productivity Tradeoffs for GPU Kernel Programming on Blackwell Architecture

Device Virtual Machine (DVM)

MegaTrain: Full Precision Training of 100B+ Parameter Large Language Models on a Single GPU

Agentic Code Optimization via Compiler-LLM Cooperation

Most viewed papers (last 30 days)