A hands-on journey into CUDA C/C++ and GPU programming, focused on understanding how NVIDIA GPUs execute parallel workloads and how GPU computation can be optimized for real-world AI and computer vision systems.
- Understand GPU architecture and CUDA's execution model
- Learn threads, blocks, grids, and warps
- Work with CUDA memory hierarchies
- Implement parallel algorithms using CUDA C/C++
- Understand synchronization and concurrency
- Explore CUDA streams and asynchronous execution
- Profile and optimize GPU workloads
- Understand CUDA integration with PyTorch
- Build the foundation required for deploying AI/CV workloads on NVIDIA edge hardware