Open-source alternatives to HVM2

xtensor is a C++ multidimensional array library for numerical computing that provides N-dimensional containers with an interface mirroring the NumPy API. It utilizes a lazy evaluation expression engine to defer numerical computations until assignment, which minimizes memory allocations and intermediate copies. The library features a foreign memory array adaptor that allows it to wrap external buffers, such as NumPy arrays, to perform numerical operations in-place without duplicating data. It further optimizes performance through lazy broadcasting and a system that manages the lifetime of temp

c3c is the compiler for the C3 programming language, transforming source code into executable binaries, static libraries, or dynamic libraries using an LLVM backend. It implements a system based on result-based error handling, scoped memory pooling, and a semantic macro system. The compiler provides first-class support for hardware-backed SIMD vectors that map directly to processor instructions and enables runtime polymorphism through interface-based dynamic dispatch. The project covers a broad set of low-level capabilities, including manual and pooled memory management, inline assembly inte

Triton is a parallel computing framework and high-level programming language designed for writing custom compute kernels. It functions as a deep learning compiler, translating complex mathematical operations into high-throughput instructions that maximize hardware utilization and memory efficiency on graphics processing units. The framework distinguishes itself through a hardware-agnostic compute abstraction that allows developers to define kernels without manual low-level tuning. It employs just-in-time compilation to generate optimized binary instructions at runtime, utilizing static data f

This project is a technical reference and documentation suite focused on the internal architecture and operational principles of the Java Virtual Machine. It provides comprehensive guides and analysis on how the virtual machine manages class loading, memory organization, and bytecode execution. The documentation distinguishes itself by providing deep dives into specific runtime mechanisms, such as the binary decoding of class files, the hierarchical delegation model for class loaders, and the precise sequence of the loading, linking, and initialization lifecycle. It also details memory reclam

This project provides a comprehensive technical guide and framework for engineering large-scale machine learning systems. It covers the full lifecycle of model development, focusing on the infrastructure and computational principles required to build, train, and serve generative AI models across distributed GPU clusters. The repository distinguishes itself by offering deep-dive tutorials and implementation strategies for complex system challenges. It emphasizes high-performance architectural primitives, such as collective communication orchestration, distributed tensor sharding, and static gr

Dask is a parallel computing framework and distributed task scheduler designed to scale Python data science workflows from single machines to large clusters. It functions as a cluster resource manager that orchestrates computational logic by representing tasks and their dependencies as directed acyclic graphs. This architecture allows the system to automate the distribution of workloads across available hardware while managing complex execution requirements. The project distinguishes itself through a lazy evaluation engine that defers data operations until they are explicitly requested, enabl

This project is a C-based heap management library designed for systems programming environments that require manual control over memory lifecycles. It provides the fundamental utilities necessary to initialize a designated region of memory and perform dynamic allocation and deallocation of memory blocks. The library utilizes a linked-list-based tracking system to manage memory segments, employing a first-fit strategy to locate available space within the heap. To maintain memory organization and minimize fragmentation, it implements boundary-tag coalescing, which automatically merges adjacent

Lean 4 is a functional programming language and interactive proof assistant used to formalize mathematics and verify software correctness. It functions as a dependent type theorem prover and a formal verification tool that allows users to construct mathematical proofs and ensure program correctness. Additionally, it serves as a logic-based source for generating verified datasets used to train and benchmark artificial intelligence reasoning systems. The system distinguishes itself through a small-kernel verification model, where all proofs are verified by a trusted core of basic logical rules.

Lazy.js is a JavaScript library that implements a lazy evaluation model for processing collections and data streams. It defers all computation until iteration begins, building chains of transformations that execute only when values are consumed, avoiding intermediate arrays and buffering. The library wraps data sources into a uniform sequence interface, enabling operations like map and filter to be chained together without materializing intermediate results. The library extends lazy processing beyond simple collections to handle asynchronous data sources, DOM events, strings, and Node.js stre

This project serves as a comprehensive educational resource for learning parallel programming and high-performance computing using graphics processing units. It provides technical guidance on the fundamental paradigms required to offload computationally intensive tasks from a host system to specialized hardware accelerators. The materials cover the core methodologies for managing data-parallel operations, including the orchestration of memory between host and device spaces and the organization of threads into structured grids and blocks. It details the execution models necessary to distribute

This repository serves as a centralized collection of state-of-the-art deep learning architectures and reference implementations designed for research and application development. It provides a comprehensive toolkit for computer vision and natural language processing, offering pre-built models and training pipelines for tasks ranging from image classification and object detection to complex sequence modeling. The project distinguishes itself by providing a flexible execution harness that manages the entire training lifecycle, including data ingestion and backpropagation. It supports scalable

This project is a cross-platform machine learning inference engine designed to execute pre-trained models across diverse operating systems and hardware environments. It functions as a standardized execution framework that manages the entire lifecycle of model inference, from loading and graph optimization to hardware-accelerated execution and generative sequence management. The runtime distinguishes itself through a highly modular architecture that decouples model logic from hardware-specific kernels. By utilizing an execution provider abstraction, it enables developers to offload computation

Goja is a JavaScript engine and ECMAScript compliant interpreter implemented entirely in Go. It serves as an embedded scripting engine that allows Go applications to execute JavaScript code and integrate a programmable scripting layer without relying on Cgo or external native dependencies. The project functions as a bridge between Go and JavaScript, enabling bidirectional data exchange and function invocation. It allows Go hosts to expose native structs, slices, and maps as JavaScript objects and arrays, while providing mechanisms to export script values and functions back into native Go type

mcp-context-forge is a Model Context Protocol federation gateway that unifies diverse AI tool servers and APIs into a single consistent interface for discovery and execution. It acts as a centralized proxy that aggregates multiple servers and APIs, allowing AI agents to access and invoke a unified set of tools, prompts, and resources. The project distinguishes itself through a multi-protocol translation bridge that converts communication between standard I/O, SSE, gRPC, and REST to enable interoperability between disparate tool servers. It includes a comprehensive LLM evaluation framework for

Cactus is an on-device AI inference engine designed for executing large language models, vision models, and speech-to-text systems on mobile and wearable hardware. It provides a programmable tensor computation graph for defining sequences of matrix operations and activation functions, alongside a local retrieval augmented generation framework that grounds model responses using local text files. The project features a multiplatform SDK with language bindings for integrating AI capabilities into mobile applications and a model conversion system that transforms external model formats for optimiz

mctx is a framework for executing high-performance tree search and state simulations to generate policy targets for neural networks. It functions as a compiled search engine and neural dynamics simulator that predicts state transitions and rewards using learned representations. The project implements a vectorised tree search capable of running parallel search operations across input batches. It utilizes a policy target generator to convert search results into action weights used for training and refining neural network policies. The system covers reinforcement learning workflows by integrati

Surge is a Swift library for high-performance numerical analysis, linear algebra, digital signal processing, and accelerated image manipulation. It utilizes the Accelerate framework to provide hardware-accelerated tools for matrix mathematics and signal processing. The library provides specialized capabilities for digital signal processing, including convolution, signal similarity analysis through cross-correlation, and domain transformations using fast Fourier transforms. It also includes a suite of tools for the rapid transformation and analysis of pixel buffers and image data. Beyond sign

Boost is a collection of portable, high-performance source libraries that extend the C++ standard library. It provides a wide range of reusable components, data structures, and algorithms designed to add capabilities to the base language across different platforms. The project is distinguished by its extensive focus on compile-time template metaprogramming and generic programming. It implements advanced architectural patterns such as policy-based design, concept-based type validation, and the use of SFINAE for conditional template resolution to minimize runtime overhead. The library covers a

This project is an educational implementation of a relational database engine written in C. It functions as a SQLite clone, demonstrating the internal mechanics of a database system through a C-based systems project that focuses on manual memory management and file I/O. The engine is distinguished by its use of a bytecode virtual machine, which executes database operations by compiling SQL statements into low-level instructions. It utilizes a B-tree database engine to organize records in a balanced tree structure, ensuring efficient insertion, search, and range scanning. The system covers co

Frege is a purely functional programming language that compiles to JVM bytecode, providing Haskell-like semantics for the Java platform. It is built around a Haskell-inspired compiler that implements non-strict evaluation and a static type inference system to ensure data immutability and prevent side effects. The project distinguishes itself through a sophisticated type system featuring rank polymorphism, type-class based dispatch, and static purity enforcement. It includes a JVM language bridge and a foreign function interface that map Java classes and interfaces into functional types, allow

FLAnimatedImage is a high-performance rendering engine and library for playing animated GIF images on iOS. It functions as an animation engine and image memory manager designed to ensure smooth frame transitions and consistent playback timing across different device environments. The project focuses on maintaining consistent animation speed through variable frame delay support and optimizes rendering efficiency to sustain high frame rates on mobile devices. It includes a memory management system to prevent application crashes and performance degradation during the playback of large or numerou

Bigcache is a thread-safe, in-memory key-value store for Go designed to minimize garbage collection pauses. It functions as a memory-optimized cache capable of storing gigabytes of data while maintaining low latency during high-volume operations. The system avoids garbage collection overhead by managing memory through large byte arrays and manual allocation. It utilizes a circular-buffer layout to reuse space and lock-striped sharding to reduce contention during concurrent read and write access. The project provides capabilities for high-volume in-memory caching and low-latency data retrieva

h2o-3 is a distributed machine learning platform and automated machine learning framework designed for training and deploying predictive models using distributed in-memory computing. It functions as a deep learning framework and a distributed model scoring engine, capable of operating as a Kubernetes ML cluster to process large datasets in parallel. The platform distinguishes itself through automated machine learning capabilities that automatically select the best algorithms and hyperparameters to optimize model performance. It provides specialized deep learning toolkits for tasks including i

fp-ts is a TypeScript library that brings pure functional programming patterns to the language through algebraic data types, type class abstractions, and composable combinators. It provides foundational data types like Option for optional values, Either for typed error handling, and Task for lazy asynchronous computations, all designed to make invalid states unrepresentable and side effects explicit. The library is built on category theory concepts, offering type classes such as Functor, Applicative, Monad, Semigroup, and Monoid with lawful instances for common data structures. The library di

cel-go is a Go implementation of the Common Expression Language, providing a type-safe engine for the evaluation of non-Turing complete, side-effect-free expressions. It functions as an embedded expression engine that converts human-readable strings into type-checked abstract syntax trees to ensure safe execution. The system utilizes a compile-once evaluate-many model, parsing and validating expressions into cached programs for repeated execution across different datasets. It further distinguishes itself as a partial state evaluator, using commutative logic to compute results and return resid

EASTL is a C++ Standard Template Library implementation consisting of containers, iterators, and algorithms. It provides cross-platform data structures and a template-based algorithm library designed for use in resource-constrained game engine environments. The library focuses on game engine memory management, providing specialized utilities that ensure predictable memory allocation and high-performance access for real-time applications. These containers maintain consistent behavior across different operating systems and hardware platforms. The project covers high-performance C++ development

This project is a comprehensive collection of frontend development best practices and standards. It provides guides for writing semantic HTML, efficient CSS, and functional JavaScript to improve overall web performance and accessibility. The repository establishes frameworks for modern CSS architecture, focusing on specificity management and the use of Flexbox and Grid for predictable layouts. It outlines patterns for functional JavaScript programming, emphasizing the use of pure functions and composition to reduce execution time and dependency bloat. The guidelines cover several core capabi