In this workshop, we will learn how to use Numba to speed up Python code. Numba is a just-in-time compiler that translates Python functions into machine code, which can significantly boost performance for numerical and scientific tasks. We will cover the basics of applying Numba’s features, optimizing functions, and integrating with libraries like NumPy. By the end, you’ll understand how to use Numba to improve your code’s efficiency and performance.

If you have any questions regarding this training, please contact training.nci@anu.edu.au.

Prerequisites

Experience with Python.
Experience with bash or similar Unix shells.
Having a valid NCI account and vp91 membership
The training session is driven on the NCI ARE service. You can find relevant documentations here: ARE User Guide.

Learning Outcomes

After this training session, you will be able to

Learn the basic concepts of parallel programming.
Parallelise Python code using Numba.
Use GPUs in Numba.

Topics Covered

CPU Parallelisation
- Working with Numba
- Just-in-Time (JIT) Compiler
- LLVM and Numba
- Eager Compilation in Numba
- Compiler Modes in Python
- No GIL mode
- Caching Compiled Functions
- Automatic Parallelisation
- Universal Functions
- Vectorize
GPU Parallelisation
- CUDA and Numba
- Memory Management
- Streams
- Vectorization in GPU
- Reduction in GPU