EPICURE (Unlocking European-level HPC Support) project, in collaboration with IT4Innovation, is organizing a hybrid course on Programming GPUs with CUDA and C/C++ on 9 and 10 October 2024. The main goal of the course is to introduce participants to programming GPU-accelerated applications using CUDA programming. The registration deadline is 2 October 2024.
In this course, presenters will explain the main principles of heterogeneous or accelerated computing, with a brief hardware description of GPU-accelerated supercomputers. The aim is to provide a proper understanding of how to design CUDA code. The course is intended for beginners in GPU programming using CUDA. It will cover topics such as parallelization with basic examples, management of data transfers between CPU and GPU memory, types of GPU memory and their usage, execution of parallel threads, and key parallel computing patterns in CUDA.
