Where will China get its compute in 2026?
Over half of the compute will likely be legally imported NVIDIA H200s, but other sources—domestic production, proxy fabrication, and smuggling—matter too, as does remote access.
Though the AI chip export controls have gaps, and can be improved, they go a long way towards reducing the amount of compute that Chinese AI companies get. If you think, as I do, that compute is of great strategic importance, and that it’s better for the US to have a comfortable lead over China, this is a good thing. The American compute advantage is probably the main reason why Chinese AI models have lagged on average 7 months behind the frontier. In this post, I make some rough estimates, using publicly available data, of how much compute China will acquire in 2026 through each of four…
related reading
- How Much Compute Does China Have?chinatalk.media
- 2028: Two scenarios for global AI leadership \ Anthropicanthropic.com
- No Jensen, Not All Compute is Created Equalchinatalk.media
- Should the US Sell Blackwell Chips to China? | IFPifp.org
- China's AI Models Are Closing the Gap—but America's Real Advantage Lies Elsewhere | RANDrand.org
- Dario Amodei — On DeepSeek and Export Controlsdarioamodei.com
- Should the US Sell Hopper Chips to China? | IFPifp.org
- China's AI Chip Landscape: A Complete Map of Who's Building Whatcrossingriver.substack.com
- Ban the H20: Competing in the Inference Agesubstack.com
- Dylan Patel — Deep dive on the 3 big bottlenecks to scaling AI computedwarkesh.com
- AI Chip Owners Documentation – Methodology | Epoch AIepoch.ai
- Preventing AI Chip Smuggling to China | CNAScnas.org