Awesome

GPU Sprite Instancing :zap:

I. Overview

• GPUSpriteInstancing is a Unity package that provides efficient GPU-based sprite rendering using compute buffers and instancing
• Optimized with Unity Burst compiler and Job System for high-performance sprite rendering

II. Features

• GPU Instanced Sprite Rendering
• Batched Drawing System
• Sprite Atlas Support
• Memory Management
• Buffer Management
• Job System Integration
• Burst Compilation

• GPU Sprite Manager
• Instance Renderer
• Custom Shader
• Sprite Data Structure

III. Setup

> Requirements

• Unity 2022.3 or higher
• Burst Package
• Mathematics Package

> Installation

1. Open Package Manager from Window > Package Manager
2. Click on the "+" button > Add package from git URL
3. Enter the following URL:

https://github.com/centhoang/GPUSpriteInstancing.git?path=/Assets/GPUSpriteInstancing

Or open Packages/manifest.json and add the following to the dependencies block:

{
    "dependencies": {
        "com.centhoang.gpu_sprite_instancing": "https://github.com/centhoang/GPUSpriteInstancing.git?path=/Assets/GPUSpriteInstancing"
    }
}

[!IMPORTANT] Use tag(#) to avoid auto-fetching undesired version

_Example

https://github.com/centhoang/GPUSpriteInstancing.git?path=/Assets/GPUSpriteInstancing#1.1.0

> Intergration

1. Add a GPUSpriteManager component to a GameObject in your scene
2. Assign the GPUSpriteInstancingMat material (or your material which suits the system) to the manager
3. Create your sprite atlas texture
4. Use GPUSpriteManager.UpdateSprites() to render your sprites

IV. Usage

> Basic Implementation

// Create sprite data
var spriteData = new NativeArray<SpriteRendererData>(count, Allocator.Temp);
// Fill sprite data with positions, UVs, etc.

// Update and render sprites
spriteManager.UpdateSprites(spriteData, atlasTexture);

> Best Practices

• Use shared texture atlases to minimize draw calls
• Keep sprite batches within the 1023 instance limit per draw call
• Dispose of NativeArrays when no longer needed
• Consider using object pooling for dynamic sprite systems

V. Performance

• Optimized for rendering hundreds of thousands of sprites
• Minimal CPU overhead using Job System
• Efficient memory usage with compute buffers
• Batched rendering to reduce draw calls