Apple Metal - Vector Addition Demo

This iOS project is the "Hello, World" of GPU compute programming. It demonstrates how to add two arrays of length N by taking advantage of GPU parallelism, where each of the N threads is responsible for adding a pair of elements.

This repository is a minimal, educational example that demonstrates how to use Apple's Metal framework for GPU programming in Swift. It's designed to help you understand the full end-to-end pipeline of compute programming with Metal in a real Swift app. It shows how to prepare data on the CPU, send it to the GPU, run a Metal kernel, and read back the results.

The code includes extensive inline comments explaining the functioning of each step and the design choices made.

Core Functionality

The project includes two core classes:

1. MetalKernelHelper.swift: All the boilerplate for setting up Metal runtime.

   It handles:

   • Metal device creation
   • Kernel loading from the default Metal library
   • Compute pipeline setup
   • Command buffer + encoder setup
   • Utility methods for creating buffers from arrays or constants

2. MetalKernelDispatcher.swift: For dispatching the vector_add kernel to the GPU for execution.

   It handles:

   • Input array preparation (two [Float] arrays)
   • Data loading into GPU buffers
   • Invoking the Metal kernel for execution
   • Timing the kernel execution
   • Reading back the results into a Swift array

Topics Covered

The projet shows how to:

• Set up a complete Metal compute pipeline in Swift
• Write and compile a Metal kernel (.metal file)
• Transfer data between CPU and GPU using Metal buffers
• Dispatch GPU workloads with dispatchThreadGroups
• Get results back from the GPU

Run Requirements

• This is an iOS project so you'd need an Apple Mac to run it
• Xcode with Metal support

License

MIT License — free to learn from and build on. A star ⭐️ would be awesome if this helped you!

Author

Created by Manish Kumar

Questions welcome!