tryriscv.org
Get a recommendation

Contribute

Report errors, suggest parts, contribute measured data, or help with the daily data update.

Three fastest ways to help

Report a data error A wrong parameter, a dead link, a mistyped part number — just include the part and a correct source. Suggest a new part The vendor page, the datasheet link, and key parameters — one issue is enough. Submit a selection request Tell us the scenario, clock speed, memory, peripherals, and power constraints; we will come back with a candidate list.

How the data is maintained

There is exactly one source of truth: data/mcu/*.yaml. The pipeline is:

Vendor website / datasheet
        │  manual or AI extraction
        ▼
data/mcu/<vendor>.yaml   ← single source of truth (with sources / confidence)
        │  python3 scripts/normalize.py
        ▼
static/data/mcus.json    ← consumed by the front-end filters and comparison
        │  hugo --minify
        ▼
GitHub Pages

normalize.py validates required fields, ID uniqueness, numeric ranges, and link formats, then generates the JSON used by the filter pages. CI enforces the checks before deployment.

Contributing one record

  1. Fork the repository and create a branch;
  2. Add a record to data/mcu/<vendor>.yaml using the fields described in research/SCHEMA.md;
  3. Provide at least one official source link and fill in confidence;
  4. Run python3 scripts/normalize.py --check and hugo --minify locally;
  5. Open a pull request describing the data source and the date it was captured.

Daily updates

The repository ships a reusable update skill and check scripts: re-crawl new vendor parts, recheck dead links, and generate a change summary. Maintainers follow docs/DAILY-UPDATE.md.