A new dataset has been developed to assess the water efficiency of data centers in 41 African countries, considering both direct cooling and indirect electricity generation water usage. The research estimates that running a 10-page report on Llama-3-70B could use 0.66 liters of water, while GPT-4 could consume up to 59 liters for the same task. These figures, based on 2024 data, highlight significant differences in water intensity for electricity generation across African nations, with many consuming less water than the global average. AI
IMPACT Highlights the significant water consumption of AI models, prompting a need for more efficient infrastructure and model development, especially in regions like Africa.
RANK_REASON The cluster contains an academic paper detailing a new dataset and research findings. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →