Computing power available to Chinese organisations for artificial intelligence is expected to almost double next year to around 8.8 million Nvidia H100-equivalent chips, up from roughly 4.6 million, but around 65 per cent of it will come from US-designed hardware, according to a report by the Washington-based Institute for Progress, the South China Morning Post reports.
The research found Chinese access to computing through remote and offshore cloud services alone is set to rise from roughly one million H100-equivalents this year to 2.8 million by the end of 2027.
The report estimates the United States holds about a 16-fold lead in AI training compute, yet leading Chinese models trail their American rivals by only about six and a half months. The findings point to continued Chinese dependence on foreign chips despite export controls.
Source: South China Morning Post
