cool-japan/scirs

SciRS2 - Scientific Computing and AI in Rust., providing SciPy-compatible APIs while leveraging Rust's performance, safety, and concurrency features. Unlike traditional scientific libraries

What it solves

SciRS2 provides a high-performance, production-ready scientific computing and machine learning infrastructure that eliminates the need for complex system dependencies. It solves the common "dependency hell" associated with scientific libraries by replacing traditional C, C++, and Fortran backends with a 100% pure Rust implementation, ensuring effortless installation and consistent behavior across all platforms.

How it works

It leverages Rust's ownership and concurrency model to provide safe, memory-efficient scientific computing. By using specialized pure-Rust crates (such as OxiBLAS for linear algebra and OxiFFT for signal processing), it achieves significant performance gains through SIMD optimization. The library is structured as a workspace of 29 specialized crates, offering a comprehensive API that mimics SciPy while maintaining native Rust safety and cross-platform compatibility (including WebAssembly).

Who it’s for

This project is for developers and researchers who need high-performance scientific computing, data analysis, or machine learning tools in Rust. It is ideal for those who want to avoid the maintenance burden of linking against external C/Fortran libraries or who require consistent, reproducible performance across Linux, macOS, Windows, and WebAssembly environments.

Highlights

  • Pure Rust: Zero C/C++/Fortran dependencies, ensuring simple builds and portability.
  • Ultra-Fast: Delivers 10-100x performance improvements via SIMD optimization.
  • Memory Safe: Built-in protection against memory leaks and data races using Rust's ownership system.
  • Battle-Tested: Includes over 38,000 tests and 3.1 million lines of code to ensure production-grade reliability.
  • Comprehensive: Covers linear algebra, statistics, machine learning, signal processing, computer vision, and symbolic mathematics.

Related

  • Project
    Cognitive-AI-Systems/pogema

    A flexible grid-based environment for Partially-Observable Multi-Agent Pathfinding (PO-MAPF), designed for testing decentralized navigation and collision avoidance in multi-agent systems.

  • Project
    oracle-samples/oci-data-science-ai-samples

    Oracle Cloud Infrastructure Data Science and AI Samples is a collection of practical demos, tutorials, and code examples for OCI Data Science and AI services. It helps data scientists and developers learn how to use features like model deployment, jobs, pipelines, and LLM support through organized notebooks and scripts.

  • Project
    mllam/neural-lam

    A repository of graph-based neural weather prediction models for Limited Area Modeling (LAM) using PyTorch and PyG.

  • Project
    openclimatefix/metnet

    A PyTorch implementation of Google Research's MetNet models for short-term and global precipitation forecasting using satellite and atmospheric data.