Reducing The Carbon Footprint Of Your Software Systems
summary
Uses the Green Software Foundation’s Software Carbon Intensity (SCI) spec and the formula E×I + M per R to produce concrete numbers: a Raspberry Pi measured at ~0.003 kWh (annualized) yields ~19 gCO2e/year; load tests plus CPU/memory coefficients and grid-intensity APIs show TigerGraph + API ≈ 118 gCO2e/hour and Neo4j ≈ 191.4 gCO2e/hour. Two simple fixes—move to a lower-grid-intensity region and downsize the VM—cut emissions ~44–45%. Provides tooling like the Carbon Aware SDK and planned CarbonQL API to automate these measurements.
AI-written summary · about
chapters (10)
- 0:00Introduction to green software
- 3:27Principles of green software
- 7:42Handprint vs footprint
- 10:30Green software fundamentals
- 13:37The SCI specification
- 15:46Calculating SCI scores
- 18:40Measurement techniques
- 26:42Practical case studies
- 42:49Open source developer tools
- 52:02Optimization and certification
related talks
34mCarbon Footprint–Aware Software Development
20mCo2 Emission Efficiency Of Software Platforms: Is There A Price For Sustainability?
43mGreen Software – How To Make Our Applications Carbon Aware?
38mGreen software: How can tech contribute to worldwide sustainability effort?
14mSLOs, SREs, and GHGs
37mHow to measure your software's impact on the planet
41mWriting Greener Software Even When You Are Stuck On-Prem
45mLaying the Foundation on OSS Environmental Sustainability