China Puts Standards Behind Embodied AI Data
A drafting group forms as the country reports 3.1 million PFLOPS of total compute, second only to the United States.
China’s national data standardisation committee held its second standards week in Xiamen on 10 November 2026, formally establishing a drafting group for embodied AI data standards alongside four sectoral groups covering weather, health, power and commerce. The National Data Administration disclosed at the same event that the country’s total compute has reached 3.1 million PFLOPS, ranking second worldwide. Two caveats matter: that figure covers total compute across basic, intelligent and supercomputing capacity rather than intelligent compute alone, which stood at 1.59 million PFLOPS at the end of 2025; and the shortage holding back embodied AI is reusable standards, not raw volume.
