AN EVEN EASIER INTRODUCTION TO CUDA
Accompanying Mark Harris’s renowned blog post, “An Even Easier Introduction to CUDA,” this notebook guides you through the fundamentals of crafting CUDA kernels for NVIDIA GPUs in a massively parallel manner. By completing this notebook, you’ll gain the ability to:
– Launch CUDA kernels massively in parallel on NVIDIA GPUs.
– Organize parallel thread execution for extensive dataset sizes.
– Manage memory interchange between the CPU and GPU.
– Profile your CUDA code for performance enhancements.
Upon finishing, you’ll be well-versed in launching massively parallel CUDA kernels for processing data on NVIDIA GPUs.
While no novel coding is required, a familiarity with the following will enhance your comprehension of the material:
– Writing, compiling, and running C or C++ code.
Recommended resources for prerequisites:
– The interactive tutorial at learn-c.org.
– CUDA C++
– “Fundamentals of Accelerated Computing with CUDA C/C++“: A self-paced course tailored to novice CUDA C++ programmers.
– “Fundamentals of Accelerated Computing with CUDA Python“: A self-paced course designed for those new to CUDA Python programming.
Digital skills for all
Digital skills for the workforce
Digital skills for ICT professionals
Format of the training
Duration of the training
Type of training
Language of the training
Country providing the training