How Hungry is AI? Benchmarking Energy, Water, and Carbon Footprint of LLM Inference · HackerPapers discussions

How Hungry is AI? Benchmarking Energy, Water, and Carbon Footprint of LLM Inference — mentioned by big_toast

Tue, 14 Apr 2026 17:25:40 GMT

Page 38 has an interesting comparison saying a session of 8 "medium length" messages uses roughly 10Wh, the equivalent of ~2 phone charges. The data dashboard[1] I think is defining medium length as 1000 tokens in, 1000 tokens out. And the page before suggests GPT-5 mini (medium) used 3x the energy of GPT-4o (Aug) which I guess I assumed would be the opposite ratio. --- I guess these estimates are based on calculations for gpt-4o (Aug) from Jegham et al[2] in November 2025. [1]:https://app.powerbi.com/view?r=eyJrIjoiZjVmOTI0MmMtY2U2Mi00Z... [2]:https://arxiv.org/pdf/2505.09598

Paper and discussion history. A mention is not an endorsement.

How Hungry is AI? Benchmarking Energy, Water, and Carbon Footprint of LLM Inference — mentioned by cillian64

Thu, 21 Aug 2025 15:28:23 GMT

One random preprint I found (https://arxiv.org/html/2505.09598v2) estimates inference is 90% of the total energy usage. From some googling around, everything seems to agree that inference dominates (or is at least comparable to) training on the large commonly-used models. I was less surprised that inference dominates training after I read that chatgpt is serving billions of requests per day.

Paper and discussion history. A mention is not an endorsement.

How Hungry is AI? Benchmarking Energy, Water, and Carbon Footprint of LLM Inference — mentioned by ben_w

Fri, 15 Aug 2025 12:59:17 GMT

Says "medium" is 1,000-token. If I assume that takes 50 seconds, 40 Wh / 50 s = 2880 watts Seems a bit much? Even if I assume 1/3rd is the amortised result of the training runs and 1/6th is the cooling system, that only takes me down to the power draw of 2.4 times the H200s' 600W maximum TDP* on the mean inference run. Semi-related to main theme: why are the links so weird? For me at least, "the published findings" links to chrome-extension://efaidnbmnnnibpcajpcglclefindmkaj/https://arxiv.org/pdf/2505.09598v2 — and I opened this with Safari. Going to the corresponding arxiv.org page manually…

Paper and discussion history. A mention is not an endorsement.

How Hungry is AI? Benchmarking Energy, Water, and Carbon Footprint of LLM Inference — HN submission

Fri, 08 Aug 2025 15:24:13 GMT

How Hungry Is AI? Benchmarking Energy, Water, and Carbon Footprint of LLM Infere

1 point; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.