Learn how to use NVIDIA Nsight Graphics to profile and optimize both graphics and compute workloads, focusing on ray tracing applications. You’ll analyze GPU performance through trace data from CUDA, DirectX, and Vulkan, identifying bottlenecks at the source-code and GPU assembly levels. By mastering tools like Flame Graphs and the “top-level triage” workflow, you’ll gain the skills needed to optimize and debug shaders and compute code for maximum efficiency. Prerequisite(s):
Basic understanding of GPU programming (CUDA, Direct3D, or Vulkan). Experience with 3D graphics. Familiarity with GPU architecture and profiling/debugging tools.