- Custom occlusion culling - faster than Unity
A spline camera records 1200 identical frames for Unity Profiler and Nsight on i5-10400F and RTX 3060 at 1440p.
100% match
· <optimization><advanced><blog><mroz><benchmarks><custom><il2cpp><nsight> - Inline vs Pipeline Ray Tracing
Inline vs Pipeline Ray Tracing is a November 2025 Evolve Benchmark blog post by Manon Oomen comparing the two DirectX and Vulkan-style ray-tracing programming models as…
100% match
· <optimization><shaders><advanced><blog><comparing><gpus><hardware><nvidia> - Optimizing spatiotemporal variance-guided filtering for modern GPU architectures
Optimizing spatiotemporal variance-guided filtering for modern GPU architectures.
100% match
· <optimization><shaders><advanced><benchmarks><gpus><hardware><nvidia><passes> - Towards Robust Agentic CUDA Kernel Benchmarking, Verification, and Optimization
Towards Robust Agentic CUDA Kernel Benchmarking, Verification, and Optimization presents a Sakana AI paper introducing robust-kbench and an agentic pipeline that translates…
96% match
· <optimization><advanced><benchmarking><benchmarks><familiarity><hardware><passes><pipeline> - Optimizing GPU-driven particle system to run 4x faster
The burning-object effect already simulates and draws on the GPU with compute shaders, append buffers, indirect dispatch, and indirect instanced rendering; this piece profiles each pass and tighten…
95% match
· <optimization><shaders><blog><mroz><measured><nsight><passes><unity> - Drawing Thousands of Meshes with DrawMeshInstanced / Indirect (Toqoz)
Toqoz's November 2019 toqoz.fyi post walks through drawing very large mesh counts in Unity with Graphics.DrawMeshInstanced and Graphics.DrawMeshInstancedIndirect, with C# samples,
94% match
· <shaders><advanced><blog><context><custom><drawcalls><loop><master> - Roystan
(roystan.net) is a personal Unity blog whose articles index lists long shader tutorials plus shorter gameplay notes from 2018-2019.
92% match
· <shaders><blog><hardware><newer><passes><pipeline><scenes><srp> - SIMD Started It, SIMT Improved It
SIMD Started It, SIMT Improved It is a January 2026 ACM SIGGRAPH Blog article by Jon Peddie of Jon Peddie Research on how GPU execution models evolved from SIMD pipelines to SIMT…
90% match
· <blog><context><familiarity><gpus><hardware><newer><nvidia><readers> - Optimizing Individual Draw Calls - Game Optimization - Episode 9
[video] Optimizing Individual Draw Calls - Game Optimization - Episode 9 is an approximately 52-minute YouTube lesson by Ben Cloward in his Game Performance Optimization series.
90% match
· <optimization><captures><comparing><drawcalls><pipeline><scenes><unity> - This will rewire how you think about rendering optimization
This Procedural Pixels article by Jan Mróz presents a teaching mental model of the GPU rasterization pipeline for everyday optimization.
89% match
· <optimization><shaders><advanced><blog><mroz><deck><pdf><pipeline> - How to profile the rendering - GPU profiling basics
It separates profilers from Frame Debugger and RenderDoc, which debug resources without per-draw durations.
89% match
· <optimization><blog><mroz><drawcalls><hardware><nsight><scenes><windows> - Optimizing whole frame - case study
Every pass uses a fixed camera angle, DX12, and Nsight Graphics so frame comparisons stay apples-to-apples.
89% match
· <optimization><blog><mroz><captures><markers><nsight><unity> - Profiling the GPU in Unity Editor
The workflow targets Windows with DirectX 11 or 12 and an NVIDIA GPU (the author uses an RTX 3060).
89% match
· <optimization><blog><mroz><hardware><nvidia><unity><windows><nsight> - The performance impact of vertex shader exports
The performance impact of vertex shader exports is a September 2025 Interplay of Light blog post by Kostas Anagnostou measuring how many float4 attributes passed from vertex to…
88% match
· <optimization><shaders><advanced><blog><hardware><nvidia><passes><profilers> - I made this shader run 5x faster
The original shader layers animated cellular (Voronoi) noise with distorted UVs and procedural value noise, with no texture assets.
85% match
· <optimization><shaders><advanced><blog><mroz><measured><micro><procedural> - Texture Baking - Trading VRAM for GPU performance
The case study starts with a procedural rock material built in Shader Graph: bevel edge detection on normal maps, triplanar wood and mud, and many texture samples that leave the GPU texture unit st…
85% match
· <optimization><shaders><advanced><blog><mroz><loop><nsight><procedural> - Heterogenous AoS instance encoding for a GPU-driven renderer
Heterogenous AoS instance encoding for a GPU-driven renderer is a Substack technical note by Zino on packing mixed instance types into one GPU buffer without one bloated shared…
85% match
· <optimization><advanced><blog><context><newer><pipeline><procedural><readers> - Stateless particles - rendering without a mesh
The GPU supplies only SV_VertexID and SV_InstanceID, and the vertex shader fabricates positions, UVs, and normals for each quad or triangle strip.
83% match
· <optimization><shaders><blog><mroz><custom><nsight><pipeline><procedural> - Creating a Volumetric Ray Marcher
The post builds a march loop along the camera ray inside a mesh volume, sampling tiled 2D slice atlases as pseudo-3D density, accumulating opacity, and clipping to a bounding box while supporting o…
83% match
· <shaders><blog><custom><loop><newer><procedural><readers><reference> - How Michael Abrash doubled Quake framerate
How Michael Abrash doubled Quake framerate is a technical archaeology article by Fabien Sanglard rebuilding WinQuake with and without Abrash's assembly to test Carmack's claim…
83% match
· <optimization><blog><hardware><loop><measured><micro><readers>