metal is a library for running computational tasks (GPGPU) on Apple silicon through Apple's Metal API.
Apple's Metal API is a unified framework for performing various types of task on Apple silicon GPUs. It offers low-level, direct, detailed access to the hardware (hence, metal ) for fast and efficient processing.
The processing centers around pipelines, which consist of a function to run and an arbitrary number of arguments and buffers. The metal function is parsed into a series of operations, and the arguments and buffers of data are streamed through it in SIMD groups. (For more details on SIMD groups and best practices for writing metal functions using them, see Apple's documentation on threads and threadgroups.)
This library leverages Apple's Metal API to run computational processes in a distributed, parallel method. First, a metal function is parsed, added to a pipeline, and cached. This happens once for every metal function. Then, any number of metal buffers are created. A metal buffer is an array of arbitrary length that references items of an arbitrary type. The actual type is defined in the metal function's definition. Finally, the metal function is run with the metal buffers and any static arguments. This streams the arguments and the data in the buffers through the computational operation(s) as sequenced in the metal function.
For the full documentation and example usage, see https://pkg.go.dev/github.com%2Fgreen-aloe%2Fmetal?GOOS=darwin.
- Add objective-c tests (hopefully with the XCTest framework)
- Rework the folding process. It can be a bit hard to visualize and could be easier to use.
- Performance: profile, find leaks, tighten everything up
- Better align objective-c code with Apple standards
- We can't import the Metal library directly into go code, but see if we can translate things better (like an NSError into a go error)