Awesome GitHub RepositoriesGPU-Accelerated Compilers

JAX is a hardware-accelerated array library and automatic differentiation system for numerical computing. It provides a framework compatible with NumPy that extends array operations with a just-in-time compiler to transform Python functions into optimized kernels for execution on GPU and TPU accelerators. The system differentiates itself through the use of an XLA-based compiler and a single program multiple data sharding model. These capabilities allow the library to distribute large-scale computations across multiple hardware accelerators using both automatic parallelization and manual shard

Genesis is an embodied AI simulation platform and parallelized robotics simulator designed for training general-purpose robotic agents. It integrates a physics engine for robotics that calculates collisions and movements for rigid bodies, soft tissues, and fluids, alongside a photorealistic 3D rendering engine. The platform features a domain randomization framework to vary environment parameters across parallel simulations, aiding in sim-to-real transfer. It supports the integration of real-world captured light fields and Gaussian splatting to provide photorealistic backgrounds within simulat

Genesis World is an embodied AI simulation platform designed for training robotic agents through physics-based interactions. It centers on a multi-physics simulation engine that integrates rigid body, particle, and finite element method dynamics, supported by a parallel simulation kernel compiler that translates Python functions into optimized GPU and CPU kernels. The platform features a photorealistic robot renderer that utilizes path-tracing and Gaussian Splatting to generate synthetic training data. It includes a domain randomization framework to vary lighting and physical parameters acros

Bend is a high-level parallel programming language and compiler designed to execute code across multi-core CPUs and GPUs automatically. By translating functional source code into a graph-based intermediate representation, it enables massive parallel execution without requiring manual management of threads, locks, or atomic operations. The runtime operates as an interaction net engine, where computations are represented as networks of nodes that reduce through local rewriting rules. This model utilizes a work-stealing scheduler to distribute tasks across thousands of hardware threads, ensuring

Blender is a professional 3D creation suite designed for modeling, animation, rendering, and video editing. It functions as an open-source 3D engine that provides a comprehensive framework for procedural geometry, physics simulation, and high-quality visual output. The platform is built upon a foundational architecture that utilizes data-block-based memory management and a dependency-graph-based evaluation system to handle complex scene transformations and geometry updates. The software distinguishes itself through a highly modular, node-based procedural architecture that allows users to cons

TileLang is a Python-embedded domain-specific language compiler that JIT-compiles and autotunes GPU kernels. It uses a tile-based DSL, automatic software pipelining, and parallel autotuning to generate optimized GPU kernels at runtime. It supports tensor core operations with Pythonic syntax, automatic memory management, and thread mapping. The compiler searches over tile sizes, thread counts, and scheduling policies, compiling and benchmarking candidates in parallel to find the fastest kernel. It also caches compiled binaries and tuning results to disk for reuse across sessions. TileLang inc

gpu.cpp is a lightweight C++ library for executing low-level general-purpose GPU computation across different hardware vendors and operating systems. It functions as a portable GPU wrapper, kernel orchestrator, and tensor management system using the WebGPU specification to abstract device initialization, buffer transfers, and compute shader dispatching. The library provides a framework for defining compute kernels from shader code and managing their asynchronous dispatch and synchronization. It enables the execution of cross-platform compute shaders and the orchestration of GPU tasks through

zml is a machine learning model compiler and cross-platform inference engine that transforms model descriptions into optimized executable binaries for specific hardware accelerators. It functions as a model deployment toolkit and hardware-agnostic orchestrator, utilizing a tensor-based architecture definition to provide strong type checking during the compilation process. The project distinguishes itself through the ability to shard tensors and distribute large-scale AI workloads across a logical mesh of multiple devices. It further supports the remote model lifecycle by authenticating and do

The project is a reusable collection of modular compiler and toolchain technologies designed for building optimizers, code generators, and multi-language programming environments. It provides foundational compiler frontend technologies that translate source code written in C, C++, and Objective-C into a low-level programming language and intermediate code format. This intermediate representation enables cross-language analysis, program transformation, and target-independent optimization alongside a cross-platform programming framework that allows developers to write single-source accelerated a

Tiramisu is a polyhedral C++ compiler framework designed to express and optimize data-parallel algorithms for diverse hardware accelerators. It provides a programming interface that allows developers to define mathematical expressions, manage loop iteration spaces, and organize functions targeting heterogeneous architectures. The system features an advanced compilation infrastructure that abstracts computations into a hardware-agnostic intermediate representation before lowering them into native machine code or hardware configuration bitstreams. It calculates exact data flow dependencies thro