GPU Programming

The workshop will take place at Science Park 402, 1098 XH Amsterdam. Please note that lunch and drinks at the end of the workshop are included.

From deep learning to high-performance computing, Graphics Processing Units (GPUs) are nowadays an important tool for scholars and research software engineers alike. Parallel in nature, they offer incredible computing capabilities that just a few years ago were only available in supercomputers. While using GPUs to accelerate computation becomes easier year after year, obtaining high performance from these devices still requires some knowledge of how they work and the programming model on which they are based.

In this workshop we will provide the learners with the fundamental knowledge that they need to start their journey into the world of programming GPUs. After a brief introduction to the specificities of GPUs, and how they differ from traditional processors, participants will experience various ways of using them with Python. They will get familiar with libraries such as CuPy and Numba to accelerate Python code, and have first-hand experience in writing small CUDA programs that can run directly on the GPU.

The workshop is based on the teaching style of the Carpentries, and learners will follow along while the instructors write the code on screen. More information can be found on the workshop website (will be activated once registration is live, usually 4 weeks before the start of the workshop).

Please be advised that by signing up, you agree to our Cancellation and No-Show Policy, which states that cancellations made less than 2 workings days prior to the event will incur a no-show fee. Please read the full policy here for more details.

If you won’t be able to attend, please cancel your registration (by following the instructions here) so that we can offer your seat to someone on the waiting list.

Audience

The workshop is aimed at PhD candidates and other researchers or research software engineers. We offer tickets for researchers who are affiliated with Dutch research institutions. We also offer industry tickets for attendees who are not affiliated with Dutch research institutions. We do not accept registrations by Master students.

Ticket prices

Ticket prices are as follows:

• For participants affiliated with Dutch research institutions: €225.00
• For participants from industry: €675.00

Prerequisite knowledge

The participant should:

• be familiar with Python
• be comfortable working in Jupyter
• have the ability to read and understand C code

Preferred:

• knowledge of NumPy
• familiarity with high-performance computing concepts

Syllabus

• accelerate NumPy primitives with CuPy
• accelerate Python code with Numba
• learn the basics of CUDA
• use CuPy to execute CUDA code in Python
• understand the different CUDA memories
• thread synchronization in CUDA
• overlapping computation and communication

Host or Publisher Netherlands eScience Center