Claude Opus 4.6 wrote a dependency-free C compiler in Rust, with backends targeting x86 (64- and 32-bit), ARM, and RISC-V, capable of compiling a booting Linux kernel.

Anthropic's original performance take-home, now open for you to try!

CUDA Tile IR is an MLIR-based intermediate representation and compiler infrastructure for CUDA kernel optimization, focusing on tile-based computation patterns and optimizations targeting NVIDIA te…

MLIR-based partitioning system

Low Level Software Documentation

Simple MPI implementation for prototyping or learning

IREE's PyTorch Frontend, based on Torch Dynamo.

Frontend integration for PyTorch with tt-mlir

The TT-Forge ONNX is a graph compiler designed to optimize and transform computational graphs for deep learning models, enhancing their performance and efficiency.

COCONUT-SVSM

Retargetable ML compilers for the twenty-first century!

Tenstorrent MLIR compiler

A minimal GPU design in Verilog to learn how GPUs work from the ground up

Inference Llama 2 in one file of pure C

LLM inference in C/C++

LLM training in simple, raw C/CUDA

A list of awesome compiler projects and papers for tensor computation and deep learning.

Enabling PyTorch on XLA Devices (e.g. Google TPU)

LLMPerf is a library for validating and benchmarking LLMs

🤘 TT-NN operator library, and TT-Metalium low level kernel programming model.

Shared Middle-Layer for Triton Compilation

A compiler for homomorphic encryption

The Modular Platform (includes MAX & Mojo)

Backward compatible ML compute opset inspired by HLO/MHLO

Library for specialized dense and sparse matrix operations, and deep learning primitives.

TPP experimentation on MLIR for linear algebra