Training

Parallel programming and GPU programming in Python

In this training basic concepts of programming in Python and general parallel programming are introduced. You will also work in an interactive way with important theory about graphics processing unit (GPU) computing.

Verschillende iconen met focus op icoon training
03 — 04 Feb 2020
Time
9 a.m. to 5.30 p.m.
Location
SURF Amsterdam

Python for scientific computing use

This training is organized in collaboration with PRACE: partner in advanced computing. The Python programming language has become increasingly popular among researchers because of the simplicity and availability of specific programming libraries, while the correct exploitation of heterogeneous architectures poses challenges for the development of parallel applications. To bring these two subjects together, this course focuses on the use of Python on CPU and GPU platforms for scientific computing in general.

Setup

The basic concepts of good programming in Python and general parallel programming are introduced. You will then hear more about essential theoretical concepts of GPU computing in hands-on sessions. For the exercises you make use of supercomputing facilities from SURFsara using Python with different programming libraries:

numba
PyCUDA
mpi4py

Requirements

Bring your own laptop with an SSH client for the hands-on sessions. The most important requirements for following the course are some familiarity with bash assignments and Python programming language.

More information

Read more