US labs spend 10x more on compute than China, but China builds faster and has easier energy access
Jsevillamol · x · 2026-08-18
Epoch AI researcher Campbell Hutcheson highlights the differences in AI compute between the US and China:
- Spending Gap: US labs spend approximately 10 times as much as Chinese labs on compute.
- Construction Speed: China is exceptionally fast at building out infrastructure.
- Energy Access: It is typically easier to access energy in China than in the US.
- Funding Source: Leading American companies are hyperscalers that recognized the alignment of AI with their businesses early and began pouring money into it.
More from Infra
- Brex benchmark: Over half of top-growing startups build AI infra — mattturck · 2026-08-18
- Falcata: CUDA-native GBDT rebuilds training loop, 14x faster than LightGBM — srchvrs · 2026-08-18
- Brex Data: 14 of Top 25 Fastest-Growing Vendors Are AI Infrastructure — AccBalanced · 2026-08-18
- NVIDIA Boosts Local AI with Unsloth Integration and llama.cpp Optimizations — danielhanchen · 2026-08-18
- Text Watermark Detection Does Not Require Rerunning the LLM — rasbt · 2026-08-18
- Practical guide: Recommended quantization for various AI models — Memestonks2020 · 2026-08-18