Infographic: why AI is an energy problem, from computation and cooling to the grid
CurieuxExplorer · x · 2026-10-24
An infographic breaks down why AI consumes so much power:
- Computation has a physical cost: neural networks require billions of matrix multiplications; every chat message hides vast infrastructure.
- Training vs inference: training concentrates demand in data centers; inference looks cheap per request but scales to millions of requests per model worldwide.
- Moving data costs energy too: memory↔accelerator↔network traffic is a major consumer; efficient architectures cut unnecessary movement.
- Heat and the grid: compute converts electricity into heat, requiring a second infrastructure (chillers, liquid cooling); environmental impact depends on energy mix (solar, wind, hydro, nuclear, gas, storage).
- Efficiency race: better chips, smaller models, quantization, cooling and software optimization all lower energy per task.
Takeaway: AI's energy footprint is a systems problem — the answer isn't just more power.
More from Infra
- ik_llama.cpp Fork Update: Matches Strata Speeds on a 6GB RTX 2060 — IceFog72 · 2026-10-25
- Watch live: YC Paper Club on the LLM inference stack — ycombinator · 2026-10-25
- YC Paper Club on inference: every training FLOP costs ~6x per year in serving, a $220B stack — ycombinator · 2026-10-25
- Running DeepSeek at 10 tok/s on an RTX 5090 costs more in electricity than API calls — DeMischi · 2026-10-25
- Next frontier: un-nerfed instruction sets would make AMD 'unstoppable' — TheZachMueller · 2026-10-25
- NVIDIA un-nerfs consumer cards: 'you bought the card and get the whole card' — TheZachMueller · 2026-10-25