Benchmark working group measuring and reporting power and energy for MLPerf submissions.
- Pricing
- Custom quote
- Category
- AI Governance for Sustainability
- Compliance
- Not stated
Verified 2026-10-22Website
AI Governance for Sustainability
SCI for AI (Green Software Foundation)
Ratified standard for measuring the carbon emissions of AI systems across their lifecycle.
Verified 2026-10-22
Overview
SCI for AI is a ratified standard from the Green Software Foundation's Software Standards Working Group that extends the ISO-certified Software Carbon Intensity (SCI) methodology — ISO/IEC 21031:2024 — to artificial intelligence systems. The project states that AI systems have become increasingly resource-intensive with no consistent way to measure their environmental impact, and that SCI for AI provides a consensus-based standard designed to make AI's carbon footprint transparent, comparable and actionable. GSF also publishes a whitepaper mapping SCI for AI to the EU AI Act's energy measurement requirements and a certification programme for SCI calculations.
Use cases
- Measure AI system carbon emissions over the lifecycle
- Report AI energy use against a standardised methodology
- Prepare environmental disclosures for the EU AI Act
- Certify an SCI calculation
Sources
More in AI Governance for Sustainability
Category page →Calculator and reporting template for the carbon emissions of machine learning training.
- Pricing
- Free
- Category
- AI Energy Monitoring
- Compliance
- Not stated
Verified 2026-10-22Website
Open leaderboard that rates the energy efficiency of AI models for inference.
- Pricing
- Free
- Category
- AI Energy Monitoring
- Compliance
- Not stated
Verified 2026-10-22Website