sebbbi

In shader programming, you often run into a problem where you want to iterate an array in memory over all pixels in a compute shader
group (tile). Tiled deferred lighting is the most common case. 8x8 tile loops over a light list culled for that tile. 

Simplified HLSL code looks like this:

Buffer<float4> lightDatas;
Texture2D<uint2> lightStartCounts;
RWTexture2D<float4> output;

[numthreads(8, 8, 1)]

xoofx

// |       Method |      Mean |     Error |    StdDev |
// |------------- |----------:|----------:|----------:|
// |        Calli | 0.6718 ns | 0.0013 ns | 0.0012 ns |
// |     Delegate | 1.1366 ns | 0.0099 ns | 0.0088 ns |
// | FastDelegate | 1.6239 ns | 0.0103 ns | 0.0097 ns |

// MyClassLib.cs is compiled in another project (havent tested if compiling with Fody is working with BenchmarkDotnet in the same project)
// This file is referencing BenchDelegates.MyClassLib

using System;

silvesthu

GPU Optimization for GameDev

Graphics Pipeline / GPU Architecture Overview

• 2011 - A trip through the Graphics Pipeline 2011
• 2015 - Life of a triangle - NVIDIA's logical pipeline
• 2015 - Render Hell 2.0
• 2016 - How bad are small triangles on GPU and why?
• 2017 - GPU Performance for Game Artists
• 2019 - Understanding the anatomy of GPUs using Pokémon
• 2020 - GPU ARCHITECTURE RESOURCES

phi-lira

using UnityEngine;
using UnityEngine.Rendering;
using UnityEngine.Rendering.LWRP;

public class CustomRenderPassFeature : ScriptableRendererFeature
{
    class CustomRenderPass : ScriptableRenderPass
    {
        // This method is called before executing the render pass. 
        // It can be used to configure render targets and their clear state. Also to create temporary render target textures.

adamnemecek

For background and further references see: Entity Component Systems on Wikipedia

ECS by Scott Bilas (GDC 2002)

Entity->Components->Update

• entity = class: no logic + no data OR at most small set of frequently used data (ie position)
• component = class: logic + data

foreach entity in allEntities do
    foreach component in entity.components do

jlreymendez

using System;
using Svelto.Common;
using Svelto.DataStructures;

namespace Svelto.ECS
{
    public static class ExclusiveBuildGroupExtensions
    {
        internal static FasterDictionary<ExclusiveGroupStruct, FasterDictionary<RefWrapper<Type>, ExclusiveBuildGroup>>
            _removeTransitions =

bgolus

Shader "WorldNormalFromDepthTexture"
{
    Properties {
        [KeywordEnum(3 Tap, 4 Tap, Improved, Accurate)] _ReconstructionMethod ("Normal Reconstruction Method", Float) = 0
    }

    SubShader
    {
        Tags { "RenderType"="Transparent" "Queue"="Transparent" }
        LOD 100

atyuwen

//==============================================================================================================================
// An optimized AMD FSR 1.0 implementation for Mobiles
// EASU and RCAS are combined in a single pass.
// Based on https://github.com/GPUOpen-Effects/FidelityFX-FSR/blob/master/ffx-fsr/ffx_fsr1.h
// Details can be found: https://atyuwen.github.io/posts/optimizing-fsr/
// Distributed under the MIT License. Copyright (c) 2021 atyuwen.
// -- FsrEasuSampleH should be implemented by calling shader, like following:
//    AH3 FsrEasuSampleH(AF2 p) { return MyTex.SampleLevel(LinearSampler, p, 0).xyz; }
//==============================================================================================================================
void FsrMobile(

sebas77

using System.Collections.Concurrent;
using System.Runtime.InteropServices;
using BenchmarkDotNet.Attributes;
using BenchmarkDotNet.Running;

class Program
{
    static void Main(string[] args)
    {
        BenchmarkRunner.Run<Benchmarks>();

guycalledfrank

using System.Collections.Generic;
using UnityEditor;
using UnityEditor.Build;
using UnityEditor.Rendering;
using UnityEngine;
using UnityEngine.Rendering;

class ShaderBuildProcessor : IPreprocessShaders
{
    ShaderVariantCollection whitelist;