China's leading AI models rely on Nvidia chips, as the high costs of transitioning to domestic semiconductors hinder efforts toward technological self-sufficiency.
Chinese AI labs still train on NVIDIA chips; switching to domestic Ascend costs 50% more due to CUDA ecosystem lock-in. NVIDIA's technical moat persists despite geopolitical pressure—alternative stacks won't mature for years, extending the vendor's market power across regions.
The high cost of switching to local semiconductors is a major roadblock
Aug 10, 2026
01:06 pm
What's the storyChina's leading artificial intelligence (AI) models are still being trained using NVIDIA chips, as per industry insiders speaking to South China Morning Post.
The high cost of switching to local semiconductors is a major roadblock in Beijing's quest for self-sufficiency.
High transition costs are keeping AI developers in China reliant on Nvidia chips.
China's most advanced large language models are still being built on Nvidia silicon, four years after Washington first restricted…
A growing number of Chinese AI labs are experimenting with shifting earlier model training phases onto domestic chips.
Nvidia's China AI chip market share may drop from 66% to 8% by 2026 as Beijing certifies nine domestic processors for state…
Beijing has urged local firms to use homemade chips. But is China ready to turn away from Nvidia?
Beijing aims to ‘temporarily ease the training bottleneck’ in China’s AI industry, but it remains focused on tech…
