Awesome GitHub RepositoriesDistributed Computing Frameworks

Exo is a distributed inference engine designed to run machine learning models across local hardware. It functions as a network orchestration layer that automatically discovers available devices to form a unified computing cluster, allowing users to scale artificial intelligence workloads by distributing computational tasks across multiple machines. The platform distinguishes itself through its ability to manage the entire lifecycle of local models while providing a standardized gateway for external applications. By translating local model outputs into industry-standard formats, it enables exi

Ray is a distributed computing framework designed to scale Python and Java applications across clusters by abstracting task scheduling and resource management. It functions as a resource-aware execution engine that manages task dependencies, placement, and fault tolerance across networked compute nodes. At its core, the system provides a stateful actor model, allowing developers to define classes that run in dedicated processes to maintain and mutate internal state across remote method calls. The framework distinguishes itself through a robust cross-language interoperability layer, enabling f

Puter is a browser-based desktop environment and cloud-native development platform that provides a virtualized graphical workspace. It enables developers to build and deploy full-stack web applications by integrating cloud storage, authentication, and serverless backend logic directly into the browser, eliminating the need for traditional server infrastructure. The platform distinguishes itself through a unified cloud storage layer and a distributed network runtime that facilitates peer-to-peer communication and cross-origin resource fetching. It features a sophisticated cross-window orchestr

Anoma is a distributed operating system designed to abstract the complexities of blockchain networks into a unified interface for cross-chain coordination. At its core, the platform utilizes a resource-based state machine and an intent-centric execution model, where user-defined goals are processed and settled by decentralized solvers rather than through direct, manual execution. This architecture enables the creation of applications that operate across heterogeneous distributed networks while maintaining a consistent developer and user experience. The platform distinguishes itself through a

This project is a comprehensive microservices development framework designed to build scalable, resilient backend systems. It provides a production-ready runtime that integrates stability patterns directly into the service architecture, ensuring consistent performance and reliability for both web and remote procedure call services even under heavy traffic conditions. The framework centers on an interface-first development model, utilizing a domain-specific language to define service contracts that serve as the single source of truth. This approach powers an extensive code generation ecosystem

Linera is a multi-chain smart contract platform designed for horizontal scalability through a microchain-based distributed ledger. By partitioning state into independent, parallel chains that share a common validator set, the protocol enables high-performance execution of modular applications. The system utilizes a WebAssembly-based runtime to ensure secure, platform-independent execution of contract logic across the network. The platform distinguishes itself through an asynchronous messaging framework that coordinates state changes between chains by queuing messages for execution in subseque

Hyperframes is an HTML-to-video rendering engine and composition tool that transforms web layouts and CSS into encoded video files. It functions as a headless browser video pipeline and a distributed video rendering framework, allowing users to create seekable animations and programmatic motion designs using HTML, CSS, and JavaScript. The project differentiates itself as an AI agent video orchestrator, enabling the automation of video scripts and compositions through natural language prompts. It supports distributed video encoding by splitting rendering tasks across multiple serverless functi

Dapr is a distributed application runtime that provides a sidecar-based infrastructure layer for building resilient microservices and event-driven applications. By utilizing a sidecar proxy pattern, it abstracts complex infrastructure tasks into standardized, network-accessible APIs, allowing developers to focus on application logic while the runtime handles service discovery, state management, and secure communication. The platform distinguishes itself through a pluggable component architecture and language-agnostic design, enabling services written in any programming language to interact wi

This project serves as a comprehensive, community-driven directory of high-quality open-source Python libraries and tools for machine learning, data science, and artificial intelligence. It functions as a centralized resource for developers to discover, evaluate, and track the maintenance status of software packages across the entire machine learning ecosystem. The platform distinguishes itself through automated popularity tracking and data-driven content curation, which programmatically validate and rank projects based on community activity and development velocity. By organizing these tools

This project is a functional programming library and toolkit for building production TypeScript applications. It provides a system for managing concurrency, error handling, and resource lifecycles using functional effects. The project distinguishes itself through a comprehensive suite of specialized toolkits, including a dependency injection framework for decoupling service implementations, a workflow orchestrator for coordinating durable processes, and a SQL database toolkit for consistent data operations across multiple dialects. It also implements an OpenTelemetry instrumentation library f

Bullet3 is a professional physics simulation engine designed for calculating rigid body, soft body, and collision dynamics within 3D environments and robotics applications. It functions as a computational framework for determining complex geometric intersections and contact manifolds between objects in simulated space. The library distinguishes itself through a distributed rendering framework that scales heavy graphical workloads and scene generation tasks across large clusters of machines. This capability enables the production of massive datasets by distributing complex scene generation acr

Dask est un framework de calcul parallèle et un planificateur de tâches distribué conçu pour mettre à l'échelle les flux de travail de science des données Python, des machines uniques aux grands clusters. Il fonctionne comme un gestionnaire de ressources de cluster qui orchestre la logique computationnelle en représentant les tâches et leurs dépendances sous forme de graphes acycliques dirigés. Cette architecture permet au système d'automatiser la distribution des charges de travail sur le matériel disponible tout en gérant des exigences d'exécution complexes. Le projet se distingue par un moteur d'évaluation paresseuse qui diffère les opérations sur les données jusqu'à ce qu'elles soient explicitement demandées, permettant une optimisation globale du graphe et une allocation efficace des ressources. Il intègre le déversement de données conscient de la mémoire pour éviter les plantages du système lors du traitement de jeux de données dépassant la mémoire disponible, et il utilise la fusion de graphes de tâches pour combiner des séquences d'opérations en étapes d'exécution uniques, minimisant la surcharge de planification et la communication entre nœuds. La plateforme fournit une surface de capacités complète pour l'analyse de données à grande échelle, incluant le support pour l'apprentissage automatique distribué, l'intégration du calcul haute performance et le traitement de données parallèle. Elle offre des outils étendus pour la gestion du cycle de vie des clusters, le profilage des performances et la surveillance en temps réel de l'exécution des tâches. Les utilisateurs peuvent déployer ces environnements sur diverses infrastructures, incluant le matériel local, les fournisseurs cloud, les systèmes conteneurisés et les clusters de calcul haute performance.

Meshroom is a node-based photogrammetry software designed to transform collections of two-dimensional images into three-dimensional models and scene geometry. It provides a visual interface for constructing and managing modular data pipelines, allowing users to automate complex computer vision tasks such as feature extraction, depth map estimation, and mesh generation. The software distinguishes itself through a distributed computational framework that dispatches resource-intensive tasks across local hardware or remote render farms. By utilizing a directed acyclic graph execution model, it en

QuantAxis is a quantitative trading platform and algorithmic trading framework. It provides a comprehensive local environment for backtesting strategies, managing financial market data, and executing trades across stocks, futures, and options markets. The system distinguishes itself through a distributed task scheduler that spreads asynchronous computations and heavy mathematical workloads across a network of remote agents. It incorporates a multi-account trading interface to standardize the monitoring of positions and the execution of orders across various brokerage accounts. The platform c

Metaflow is a Python machine learning framework and MLOps workflow orchestrator designed to manage the lifecycle of data pipelines from local prototyping to production. It serves as a distributed compute manager and an experiment tracking system, enabling the creation of reproducible pipelines that transition between development and high-availability production environments. The framework distinguishes itself through an integrated checkpointing system that automatically persists intermediate data artifacts to remote storage, allowing failed runs to be resumed from the last successful step. It

Hyperopt is a Python library for hyperparameter optimization designed to minimize scalar-valued objective functions. It operates as a stochastic search space engine that finds optimal input parameters by searching through real-valued, discrete, and conditional spaces. The framework distinguishes itself through its support for complex search space configurations, allowing for conditional parameter hierarchies where specific hyperparameters are sampled only if their parent parameters meet certain criteria. It is built as an asynchronous optimization framework, decoupling the generation of searc

ZenML is an extensible machine learning orchestration framework designed to manage the end-to-end lifecycle of data pipelines and AI agent workflows. It functions as a durable orchestrator that executes machine learning tasks as directed acyclic graphs, ensuring that every step is containerized for consistent performance across local, cloud, and hybrid infrastructure. By decoupling pipeline code from underlying compute and storage backends, the platform allows developers to define infrastructure-agnostic stacks that remain portable across diverse environments. The project distinguishes itself

Apache Mesos est un noyau de systèmes distribués et un gestionnaire de ressources de cluster qui abstrait le CPU, la mémoire et le stockage à travers un pool de nœuds. Il fonctionne comme un orchestrateur d'infrastructure distribuée, fournissant une couche pour exécuter plusieurs frameworks d'orchestration sur un ensemble partagé de machines physiques ou virtuelles. Le système agit comme un moteur d'isolation de ressources, divisant un cluster partagé en conteneurs isolés pour exécuter diverses charges de travail simultanément. Il permet l'orchestration multi-framework, permettant à différents frameworks d'applications distribuées de partager une infrastructure unique pour maximiser l'utilisation du matériel. Le projet couvre la distribution de calcul à grande échelle et la gestion de cluster distribué. Ses capacités incluent la gestion des ressources distribuées et l'isolation de la puissance de calcul à travers plusieurs applications pour éviter les interférences et assurer des performances stables sur des serveurs partagés.

Volcano is a Kubernetes-native batch scheduler specialized for AI, machine learning, and high-performance computing workloads. It provides gang scheduling to atomically allocate resources for all tasks of a distributed job, preventing deadlocks from partial allocation, and supports hierarchical queue management for multi-tenant resource isolation with configurable quotas, borrowing, and preemption. Topology-aware placement optimizes communication-intensive workloads by modeling network hierarchy to minimize cross-switch latency. Volcano differentiates itself with automated orchestration of di

statsforecast est une bibliothèque de prévision de séries temporelles statistiques haute performance conçue pour générer des prévisions ponctuelles et des intervalles de prédiction. Elle fonctionne comme un framework de séries temporelles distribué qui utilise un moteur de prévision basé sur C et un sélecteur de modèle automatisé pour identifier et ajuster le modèle statistique optimal pour chaque série unique dans un jeu de données. Le système inclut également un détecteur d'anomalies de séries temporelles pour identifier les points de données inhabituels en comparant les valeurs observées aux intervalles de prévision probabilistes. Le projet se distingue par sa capacité à gérer la prévision parallèle à très grande échelle pour des millions de séries individuelles. Il y parvient grâce à un framework de calcul distribué, une exécution parallèle multi-cœur et des noyaux C compilés qui accélèrent la logique de base ARIMA et de lissage exponentiel. Le système optimise davantage le traitement à grande échelle en utilisant une disposition de données au format long et un pipeline de données à évaluation paresseuse (lazy-evaluation) pour réduire la surcharge mémoire. La bibliothèque fournit une suite complète de modèles, notamment AutoARIMA, diverses méthodes de lissage exponentiel pour la demande intermittente ou saisonnière, la décomposition Theta et la modélisation de volatilité GARCH pour le risque financier. Elle couvre des domaines de capacités plus larges tels que la prévision multivariée avec des variables exogènes, la décomposition de séries temporelles et l'évaluation de modèles via la validation croisée historique et l'analyse par fenêtre glissante. La bibliothèque s'intègre avec des structures de données haute performance comme Polars et fournit des utilitaires pour servir les modèles enregistrés en tant qu'endpoints REST pour des prédictions accessibles par réseau.