We are pleased to share that ENCCS is organizing a three-day online workshop titled GPU Programming: When, Why, and How? from 25 to 27 November 2025. This training is designed for researchers, engineers, and developers who already work with CPU-based applications and want to understand how to accelerate them using modern GPU-computing techniques. This workshop is a follow-up event to the webinar Introduction to GPU Programming announced last week.
The workshop will guide participants through the key concepts and practical strategies for GPU programming. Across three half-day sessions, attendees will explore a wide range of GPU-programming models — from directive-based approaches such as OpenACC and OpenMP to vendor-specific tools like CUDA and HIP, as well as portable frameworks including SYCL, Kokkos, and alpaka. The program also offers an introduction to GPU support in high-level languages like Python and Julia.
In addition to theory, the training includes demonstrations, hands-on exercises, and real examples to help participants evaluate GPU programming models, port existing code, and understand performance trade-offs across hardware platforms. A dedicated session on multi-GPU programming and preparing CPU applications for GPU acceleration will further equip participants with the knowledge needed to take their projects to the next level.
Researchers and developers are invited to register, especially those involved in HPC, computational science, and data-intensive research. An official institutional email is required for participation. Don't miss this opportunity to build foundational GPU-computing skills and prepare your applications for the future of high-performance computing.
