Project 5 : Kushagra

[Kushagra-Goel]

• Repo Link
• Features Implemented
  • CPU Code
  • GPU Code
  • No extra credits attempted
• Feedback
  • As a project to motivate students to explore new technologies
    • It is an amazing project, felt good to see the outputs, got to explore a different use case of GPUs.
    • Has a lot of scope to add a lot more features as a part of the core project.
    • Serves as an excellent primer for people who see themselves working on DXR.
  • As a project to learn DXR
    • Didn't learn much as most of the GPU logic was abstracted out by DXR and errors are not descriptive enough.
    • I understand the tiny tiny modules and I also get the high level flow of Ray Tracing, but I don't feel comfortable saying I understand the inter-connection between them very well as to extend the project myself to incorporate something novel.
    • By far the biggest issue, the whole project is a single deliverable with hardly any checkpoints. Unlike previous projects which had checkpoints which were very helpful to see if work till a given point was correct or not, for this one had to do the whole thing and then spend way more effort guessing where the error might be by trial and error.