pyconza-october-2018's Introduction

PyConZA October 2018

Gordon Inggs, 2018 ([email protected])

Hello, this is a repo will all of the content from my talk on 12 October, 2018. If you were there, I hope you enjoyed the talk!

From the root of the repo, these are the contents:

.
├── code # All code notebooks used
│   ├── PyConZA_OpenCL_Talk.ipynb
│   ├── java_hash.cl
│   └── vector_sum.cl
├── LICENSE
├── resources
│   ├── reading # Some of the recommended reading
│   │   ├── Debunking100xMyth.pdf
│   │   ├── TheFreeLunchIsOver.pdf
│   │   └── ViewFromBerkeley.pdf
│   └── setup_scripts # Setup scripts - tested on AWS EC2 instances
│       ├── setup_intel_cpu.sh # Assuming Ubuntu 16.04
│       ├── setup_nvida_gpu.sh # Assuming Ubuntu 16.04
│       └── setup_xilinx_fpga.sh # Assuming AWS FPGA Developer AMI
└── slides # Slides
    └── PyConZA_OpenCL_Talk.slides.html

Getting Started

In the resources directory, there is a sub-directory setup_scripts with bash scripts that can be used to setup PyOpenCL.

The Intel CPU and GPU setup assumed Ubuntu 16.04, whilst the Xilinx FPGA setup assumes the AWS FPGA Developer AMI.

pyconza-october-2018's People

Contributors

Stargazers

Watchers

pyconza-october-2018's Issues

integration of pyopencl with jupyter.

I couldn't get the PyOpenCL working with jupyter on windows 10, 64-bit. I installed pyopencl with pip install pyopencl-2019.1.2+cl12-cp37-cp37m-win_amd64.whl

So I can test code with IDLE 3.7.4 not with jupyter. It would be nice if I could.
Any tips on having this setup? Thanks.