Graphics Programming weekly - Issue 331 - March 17th, 2024

[video] Too Many Developers Ignore These Optimisations

  • the video provides a walkthrough of the steps when optimizing a voxel renderer
  • each step is explained and visualized to make it very easy to understand
  • covers a large number of optimizations and, for each, discusses the effect on performance and memory usage

Unlock Seamless Material Interchange for Virtual Worlds with OpenUSD, MaterialX, and OpenPBR

  • the article provides an overview of the different models available to describe the appearance of a mesh using OpenUSD
  • Explain the advantages and the disadvantages of the available methods

Agility SDK 1.613.0 Available Now, Including Support for GDC 2024 Showcase Features

  • the article provides an overview of all new features available in the D3D12 Agility SDK update
  • covers work graphs, new shader model, accessible memory for CPU/GPU, as well as unifications on shader creation

D3D12 Work Graphs

  • an overview from Microsft into what work graphs are
  • what possibility it enables, a brief summary of the spec
  • additionally points out the changes since the preview version last year

Work Graphs in Direct3D 12: A Case Study of Deferred Shading

  • the article provides an overview of how a G-Buffer classification system for BRDF selection can be authored using work graphs
  • provides performance numbers for the work graph and an Uber shader approach
  • discusses lessons learned, pitfalls, and performance considerations
  • It additionally discusses the open challenges and future developments

GDC 2024: Work graphs, mesh shaders, FidelityFX™, dev tools, CPU optimization, and more.

  • AMD provides a list of talks they will be presenting at GDC (starting today)
  • each talk is given a small summary and where it’s taking place

Font- and vector-art rendering with mesh shaders

  • the article provides a detailed walkthrough of how to implement Loop’s and Blinn’s font rendering approach from Siggraph 2005
  • provides an overview of the approach with many visualizations
  • shows how to implement the algorithm using D3D12 mesh shaders
  • source code is provided

Random walk method for quaternions

  • the blog post discusses a method for generating uniformly distributed quaternions
  • explains the theory of quaternion and the proposed method to generate random rotations
  • provides an implementation without using trigonometric in C++ SIMD as well as GLSL

Cross-Stage Shader Optimization

  • the whitepaper presents which tools and optimizations passes in spirv-cross and glslang are available to optimize shaders, taking advantage of knowledge from multiple shader stages
  • shows an example of how vertex outputs can be eliminated if the connected pixel shader doesn’t read the attribute

Recreating Nanite: LOD generation - faster, better, simpler

  • second part in an article series about creating a LODing system that allows per-cluster LOD selection
  • focuses on performance and quality fixes for the proposed solution from part 1

[video] 3D Gaussian Splatting! - Computerphile

  • the video provides an explanation of Gaussian Splatting and how it connects to point clouds and nerfs
  • presents the base idea and optimizations that have been applied to speed up the technique

PIX 2403.08 – New D3D12 features + misc PIX improvements

  • the blog post provides an overview of new features and improvements in PIX
  • additionally presents at which state of support the new D3D12 features, such as shader graph, shader model 6.8, etc.

Thanks to Erika for support of this series.

Would you like to see your name here too? Become a Patreon of this series.