high performance computing on graphics processing units: hgpu.org

Tal Kadosh, Niranjan Hasabnis, Vy A. Vo, Nadav Schneider, Neva Krien, Abdul Wasay, Nesreen Ahmed, Ted Willke, Guy Tamir, Yuval Pinter, Timothy Mattson, Gal Oren

View

Download (PDF)

Source codes

Tags: Code generation, Computer science, Deep learning, HPC, nVidia, nVidia A40, nVidia H100, OpenMP, Package

September 6, 2023 by hgpu

Leveraging Memory Copy Overlap for Efficient Sparse Matrix-Vector Multiplication on GPUs

Guangsen Zeng, Yi Zou

View

Download (PDF)

Tags: Computer science, CUDA, Machine learning, nVidia, nVidia A40, Sparse matrix

September 6, 2023 by hgpu

gpu_tracker: Context manager and CLI that tracks the computational-resource-usage of a code block or shell command, particularly the GPU usage

gpu_tracker: Python package for tracking and profiling GPU utilization in both desktop and high-performance computing environments

CIFAR-10 Airbench: 94% on CIFAR-10 in 3.29 second

94% on CIFAR-10 in 3.29 Seconds on a Single GPU

See all packages

* * *

high performance computing on graphics processing units: hgpu.org

Applications

Evaluating the performance portability of SYCL across CPUs and GPUs on bandwidth-bound applications

Comparing Performance and Portability between CUDA and SYCL for Protein Database Search on NVIDIA, AMD, and Intel GPUs

Compressed Real Numbers for AI: a case-study using a RISC-V CPU

Compiler-assisted distribution of OpenMP code for improved scalability

Improving the Efficiency of OpenCL Kernels through Pipes

Comparing Llama-2 and GPT-3 LLMs for HPC kernels generation

Many Cores, Many Models: GPU Programming Model vs. Vendor Compatibility Overview

host device — Generic programming in Cuda

Unified Shared Memory: Friend or Foe?

PoCL-R: An Open Standard Based Offloading Layer for Heterogeneous Multi-Access Edge Computing with Server Side Scalability

Scope is all you need: Transforming LLMs for HPC Code

Leveraging Memory Copy Overlap for Efficient Sparse Matrix-Vector Multiplication on GPUs

Recent source codes

CuPBoP-AMD: Extending CUDA to AMD Platforms

Adopter: Automated Deep Learning Optimization via DSL-based Source Code Transformation

ROCm's implementation of Gromacs

Code examples for paper on SYCL backend of Kokkos - IWOCL 2024

SimSYCL: Synchronous, single-threaded, library-only SYCL implementation for debugging and verification

GPU plugin for PySCF

QArray

Celerity: High-level C++ for Accelerator Clusters

gpu_tracker: Context manager and CLI that tracks the computational-resource-usage of a code block or shell command, particularly the GPU usage

CIFAR-10 Airbench: 94% on CIFAR-10 in 3.29 second

Most viewed papers (last 30 days)