YiOS

An operating system for RISC-V computing platforms, the OS foundation of the Hengyi Era software ecosystem

  • Ready to use out of the box

    Built-in office suite compatible with mainstream Office file formats; includes the Chromium browser, VLC player, terminal, image manager, and software center.

  • Full-stack ecosystem

    Covers five categories of key components: foundational toolchains, cloud native, databases and vector databases, middleware and big data, and AI software stacks.

  • Developer-friendly

    A complete development toolchain, with GCC, LLVM/Clang, Go, Rust, Python, and CMake all supported; build logs are fully accessible, and secondary development is supported.

Key adaptation updates: stay tuned Get adaptation information

YiOS × Computing Chips: Full-Stack Collaborative Architecture

Application Layer

Agent Applications / Cloud Computing Applications / Industry Data Center Workloads

YiMig

An intelligent migration system developed in-house based on open-source large models,

turning migration into a reusable, auditable, and rollbackable automated pipeline.

Make Every Migration a Contribution to the Ecosystem

Supports On-Premises Deployment

Product Architecture

Self-Evolving Knowledge Base · Multi-Agent Collaboration · Large Models Called as Needed · Engineering Verification

Input

Source Code Repository · Build Information · RISC-V Target Environment · Acceptance Requirements

Intelligent Migration Pipeline

  1. 1
    Engineering Analysis Agent

    Source Code Integration · Build Detection · Dependency and Toolchain Identification

  2. 2
    Cross-Compilation Agent

    riscv64-linux-gnu-gcc / g++ cross-compilation

  3. 3
    Runtime Verification Agent

    Artifact Collection · QEMU / Target Environment Execution
    Run agreed-upon tests and output verification results and evidence

YiMig Core Engine

Issue Diagnosis Agent

Analyze build and runtime logs
Extract context and generate issue fingerprints

YiMig Knowledge Base

Triggered when the knowledge base finds a match

Millisecond-Level Local Matching
No Large Model Calls
Reasoning and Repair Agent

Triggered when the knowledge base finds no match

Call a Large Language Model (LLM) for reasoning
4
Structured Deliverables

Target Binary · Runtime Dependencies · Fix Patches
Migration Report · User Guide · Verification Records

YiMig Core Engine

  • 01

    Issue Diagnosis Agent

    Analyzes build and runtime logs, extracts error signatures, and normalizes similar issues into standard "issue fingerprints", the unified key for knowledge base matching.

  • 02

    Self-Evolving Knowledge Base

    Matches issue fingerprints against verified rules in milliseconds. On a hit, the fix is output directly, with no large-model call.

  • 03

    Reasoning and Repair Agent

    On a miss, an LLM uses source code, logs and target-environment context to generate candidate fixes, which go back through the unified build and verification flow. Verified fixes are versioned into the knowledge base, so the hit rate keeps rising.

  • 80% +

    Internal knowledge base hit rate

  • Hours

    Typical project migration timelines reduced from days/weeks to hours

Why Choose YiMig?