Hello Codeforces! CUDA is the software that has contributed most to NVIDIA GPU's success. It's a core technology for AI, deep learning, and high-performance computing, but the learning curve is steep and there aren't many resources available to study it. Most importantly, it's nearly impossible to practice without a GPU. After much consideration about how to study and practice CUDA without a GPU, I created learning materials and developed a practice environment, and I'm excited to share them with the community.
Study Materials I've organized CUDA learning materials on GitHub. Step-by-step curriculum from basic syntax to advanced topics like shared memory and atomic operations GitHub: https://github.com/SungHwanYun/cudaforces
Practice Environment At cudaforces.com, you can write and run CUDA code without a GPU. Code and execute directly in your web browser Custom transpiler emulates CUDA code on CPU Website: https://cudaforces.com
Recommended For Those interested in GPU/parallel programming Students taking parallel computing courses Anyone curious about how AI/deep learning frameworks work internally
I hope this helps those who have been hesitant to study CUDA due to lack of GPU access. Questions and feedback are welcome. Thank you!
