Graphics Programming Weekly - Issue 397


Real-Time GPU Tree Generation

  • the paper presents a work graph-based procedural system for generating and rendering trees entirely on the GPU
  • shows how the model used to generate the tree structure
  • presents how this is mapped onto the work graph execution logic
  • additionally discusses how seasonal effects and wind effects are integrated


Real-Time Importance Deep Shadows Maps with Hardware Ray Tracing

  • introduces Importance Deep Shadow Maps (IDSM), a real-time algorithm for rendering shadows of semi-transparent objects
  • adaptively distributes shadow samples based on camera importance
  • achieves significant speedups over hardware ray tracing while maintaining visual fidelity


CUDA Ray Tracing 2x Faster Than RTX: My CUDA Ray Tracing Journey

  • an article discusses optimizing a CUDA-based ray tracer to outperform Vulkan/RTX implementations on the same hardware
  • demonstrates how aggressive inlining, branchless material evaluation, and custom RNGs yield significant speedups
  • provides benchmarks and practical advice for optimizing ray tracing performance
  • additionally, it presents a discussion of how to interpret the results


Performance Profiling on AMD GPUs – Part 1: Foundations

  • provides an introduction to profiling and optimizing applications on AMD GPUs for ROCm workloads
  • outlines key tools in the ROCm ecosystem for tracing, counter collection, and kernel analysis


Blackwell: Nvidia’s Massive GPU

  • provides an in-depth analysis of Nvidia’s Blackwell GPU architecture
  • compares Blackwell’s design, memory subsystem, and compute performance to AMD’s RDNA4
  • discusses architectural tradeoffs, cache design, and the implications for high-end GPUs


Intent to ship: WebGPU on Windows

  • announces Mozilla’s plan to ship WebGPU in Firefox 141 on Windows
  • notes that WebGPU is already enabled in Firefox Nightly and will expand to other platforms


Getting Started with Slang: Reflections API

  • virtual meet-up that will introduce the Slang shader reflection API
  • going to highlight common pitfalls and best practices for accessing shader metadata


[video] RE ENGINE Meshlet Rendering Pipeline

  • the very detailed presentation discusses the geometry rendering pipeline on the RE ENGINE
  • covers meshlet rendering, software rasterization, visibility buffers, shadow rendering, and more
  • additionally contains Questions and Answers documents from the audience and developers


[video] The Buffer Offset Alignment Crisis // Vulkan For Beginners #22

  • the video discusses the alignment requirements for Buffers in Vulkan
  • explains why a validation error message appears and how to resolve it
  • shows why the problem happens and how to resolve it in practice


[video] I3D 2025 Papers Session 3 - Animation & Voxels

  • video recording of I3D paper presentations
  • covering Hierarchical Neural Skinning Deformation with Self-supervised Training for Character Animation, Foveated Animations for Efficient Crowd Simulation
  • Aokana: A GPU-Driven Voxel Rendering Framework for Open World Games as well as Transform-Aware Sparse Voxel Directed Acyclic Graphs


Thanks to JanHaraldFredriksen for support of this series.


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