SLOs, SREs, and GHGs
summary
Provides a concrete method to measure software carbon: Software Carbon Intensity = energy (kWh) × carbon intensity (gCO2/kWh) + embodied carbon, normalized by requests. Uses an Azure D8asv4 VM (AMD EPYC 7452, 155W TDP, 8 vCPUs) with West Europe grid at ~265 gCO2/kWh, embodied carbon 1,560 kg over 5 years, 10 req/s → result ≈ 5.553 g CO2 per 1000 requests. Reduction levers: use less hardware, use less energy, use energy more intelligently (carbon-aware scheduling, region shifts).
AI-written summary · about
chapters (10)
- 0:00Introduction
- 0:56Energy Consumption
- 2:57Scopes
- 4:09GHG Equation
- 5:16General Equation
- 6:56Carbon Intensity
- 7:40Embodied Carbon
- 8:26Emission Rate
- 9:34Reducing Emissions
- 13:10Outro
related talks
44mCloud computing and carbon emissions
20mCo2 Emission Efficiency Of Software Platforms: Is There A Price For Sustainability?
43mGreen Software – How To Make Our Applications Carbon Aware?
45mLaying the Foundation on OSS Environmental Sustainability
57mSustainable Green Software development
41mWriting Greener Software Even When You Are Stuck On-Prem
16mGreen Code: Transforming DevOps for a Sustainable Future
59mReducing The Carbon Footprint Of Your Software Systems