https://github.com/intelpython/sample-data-parallel-extensions

Sample data parallel extensions built with oneAPI DPC++
https://github.com/intelpython/sample-data-parallel-extensions

cython dpctl gpu-computing intel-xpu oneapi pybind11 python

Last synced: about 1 month ago
JSON representation

Sample data parallel extensions built with oneAPI DPC++

• Host: GitHub
• URL: https://github.com/intelpython/sample-data-parallel-extensions
• Owner: IntelPython
• License: apache-2.0
• Created: 2022-07-08T20:44:59.000Z (over 2 years ago)
• Default Branch: main
• Last Pushed: 2024-10-21T13:31:36.000Z (2 months ago)
• Last Synced: 2024-10-21T21:48:21.592Z (2 months ago)
• Topics: cython, dpctl, gpu-computing, intel-xpu, oneapi, pybind11, python
• Language: Python
• Homepage: https://intelpython.github.io/oneAPI-for-SciPy
• Size: 45.9 KB
• Stars: 2
• Watchers: 4
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE
  • Codeowners: .github/CODEOWNERS

Awesome Lists containing this project

README

        
![Building and testing](https://github.com/IntelPython/sample-data-parallel-extensions/actions/workflows/ci.yaml/badge.svg)

# Collection of sample oneAPI Python extensions

oneAPI Python extensions is a native Python extension compiled with DPC++ and

targeting various devices programmable by DPC++, e.g. GPUs, multi-core CPUs or

accelerators such as FPGA.

This collection of examples is part of "[oneAPI for Scientific Python Community][poster]"

virtual poster by @oleksandr-pavlyk and @diptorupd presented at [SciPy 2022][scipy22] conference.

Two packages `kde_setuptools` and `kde_pybind11` implement the same Kernel Density Estimation

code executable on SYCL devices supported by DPC++, e.g. Intel CPUs, Intel GPUs.

They only differ in how they are built. One is built with plain `setuptools` while other

is build with `scikit-build` and uses DPC++ integration with CMake.

[poster]: https://intelpython.github.io/oneAPI-for-SciPy

[scipy22]: https://www.scipy2022.scipy.org/