Do you use Python for your calculations and want to take things further by harnessing the power of GPUs?
This course offers you the chance to discover GPU programming with CUDA Python and Numba, so you can speed up your calculations and develop your own CUDA kernels.
What’s on the programme:
- Understand the basics of CUDA and GPU architecture.
- Discover CUDA Python with Numba.
- Create your own custom CUDA kernels.
- Learn how to manage memory efficiently.
- Put these concepts into practice directly on the GPUs of the ROMEO2025 supercomputer
Why take part?
- Learn how to harness the power of GPUs from within Python.
- Speed up the execution of your scientific and technical applications.
Target audience:
This course is aimed at engineers, researchers and students wishing to develop or optimise their computations using GPUs.
Prerequisites:
- Proficiency in Python, particularly the NumPy library. If you do not have this, you can take free courses to get started with Python. These courses include:
- Learn the basics of the Python language – OpenClassrooms
- Get started with Python for data analysis – OpenClassrooms
- Familiarity with the basic concepts of linear algebra, including vectors, matrices and matrix multiplication.
Pre-required logistics
Prerequisites:
Proficiency in Python, particularly the NumPy library. If you do not have this, you can take free courses to get started with Python. These courses include:
Learn the basics of the Python language – OpenClassrooms
Get started with Python for data analysis – OpenClassrooms
Familiarity with the basic concepts of linear algebra, including vectors, matrices and matrix multiplication.