CUDA 4 Dummies
Organized by Austrian Scientific Computing (ASC) on 27 and 28 October 2026, this beginner-friendly, two-day online course provides a step-by-step introduction to CUDA programming, core design principles, and practical GPU workflows.
Through lectures, demos, and hands-on labs, participants will learn to extend C/C++/Fortran code to run across thousands of GPU cores in parallel and accelerate scientific computing. The course is open to anyone interested in CUDA programming, including students, researchers, and professionals in academia, industry, and public administration. No prior GPU experience is required.
CUDA is the standard API for developing code for NVIDIA GPUs, and impressive applications have emerged across fields ranging from particle physics to computational biology. It enables developers to extend standard C/C++/Fortran code with GPU-specific functions, harnessing thousands of GPU cores to run in parallel and move significant portions of computational workloads from the CPU to the device. Developers also have access to a broad ecosystem of libraries, code templates, building blocks, analysis tools, development frameworks, and an active community through the CUDA Developer Zone. For beginners, this wealth of support can feel overwhelming, making it difficult to know where to start. This course therefore offers a systematic, step-by-step introduction to GPU computing from a newcomer’s perspective. Participants will learn basic design principles, acquire core programming techniques, and explore state-of-the-art workflows that efficiently use GPUs in scientific computing.