Research map / Large language models / AI and society
Estimating the Carbon Footprint of BLOOM, a 176B Parameter Language Model
Alexandra Sasha Luccioni, Sylvain Viguier, Anne-Laure Ligozat
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| include idle consumption and embodied emissions in the footprint | BLOOM training · BLOOM-176B training | CO2eq (tonnes): 24.69 -> 50.5 | Mixed |
| train on low-carbon grid (57 gCO2eq/kWh) instead of OPT's grid (231 gCO2eq/kWh) | BLOOM · LLM training footprint | CO2eq (tonnes): 70 -> 25 tonnes | Improved |
| train on low-carbon grid instead of GPT-3's grid (429 gCO2eq/kWh) | BLOOM · LLM training footprint | CO2eq (tonnes): 502 -> 25 tonnes | Improved |
| compare raw energy of BLOOM and OPT | BLOOM · LLM training footprint | energy (MWh): 324 -> 433 MWh | Hurt |
| deploy model as real-time API with unpredictable requests | BLOOM API · BLOOM API inference | per-GPU power (W): 78-171W vs 400W TDP | Hurt |
| zero incoming requests to API instance | BLOOM API · BLOOM API inference | energy per 10 minutes (kWh): approximately 0.28 kWh | No clear effect |
| add intermediate model experiments (104B, 13B, 1B) to final model training | BLOOM · BigScience experiments | CO2eq (tonnes): 35.8 tonnes vs 24.69 | Hurt |
Closest papers by meaning
- LLMCarbon: Modeling the End-to-End Carbon Footprint of Large Language Models (ICLR 2024 · similarity 0.90)
- Holistically Evaluating the Environmental Impact of Creating Language Models (ICLR 2025 · similarity 0.89)
- CATransformers: Carbon Aware Transformers Through Joint Model-Hardware Optimization (NeurIPS 2025 · similarity 0.85)
- CO2-Meter: A Comprehensive Carbon Footprint Estimator for LLMs on Edge Devices (AAAI 2026 · similarity 0.85)
- Carbon Footprint Reduction for Sustainable Data Centers in Real-Time (AAAI 2024 · similarity 0.84)
- CarbonSense: A Multimodal Dataset and Baseline for Carbon Flux Modelling (ICLR 2025 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
