News

In a technological context that is constantly evolving, the demand from professionals for training that allows them to respond efficiently to advanced computing challenges is an increasingly present and marked reality. With this mindset in mind, EPICURE and its partner IT4I organized the “Programming GPUs with CUDA and C/C++” course. The initiative, structured and hands-on learning, aimed to empower participants to solve complex scientific problems and promote innovation in their fields.

“The main objectives of this course include teaching CUDA code development, including managing data transfers between CPU and GPU memory, efficiently utilizing different types of GPU memory, executing code in multi-GPU environment, and understanding several key parallel computing patterns,” according to Lubomir Riha, IT4I expert.

To this end, the course provides foundational knowledge and practical skills in GPU programming, which are essential for developing parallel applications for the most powerful HPC systems in the world. According to Riha, “It caters to both beginners and those with intermediate knowledge of parallel programming.”

In line with the objectives of the EPICURE project, this training aims to improve the skill set of the HPC community by providing accessible training on GPU programming, fostering innovation, and promoting the development of more efficient and robust applications in this field.