China bars ByteDance from using Nvidia chips in new data centres

• Regulators have barred ByteDance from using Nvidia chips when building new data centres.
• The move supports China’s drive for chip self-sufficiency and reduces reliance on foreign AI hardware.

What happened: ByteDance blocked from new Nvidia hardware use

Regulators in China have reportedly prohibited ByteDance, owner of TikTok, from deploying Nvidia chips in any new data centres, according to a report in the media. This comes after ByteDance became the single largest purchaser of Nvidia AI chips in China in 2025. The ban follows earlier guidance — first in August — urging local firms to stop ordering advanced Nvidia chips and instead to shift toward domestically-made alternatives.

ByteDance had been stockpiling chips to support intensive AI workloads, but the decision by regulators reflects Beijing’s effort to limit reliance on US-produced technology. Under the new rule, any fresh data-centre build-outs by ByteDance will need to use locally developed AI processors rather than Nvidia’s offerings. See also: Alejandro Estua.

The clampdown arrives amid rising geopolitical tensions, especially between the US and China over semiconductor exports, and underscores China’s ambition for technological self-reliance as export controls tighten abroad. See also: Alejandro Manzo.

Also Read: Amazon commits $15B to Indiana for AI-ready data centres
Also Read: SoftBank to invest $3 billion in Ohio factory for OpenAI data centre

Why it’s important

The ban on Nvidia chips for ByteDance signals a significant shift in China’s tech strategy — it’s not just about limiting foreign imports but actively encouraging domestic chip development. For a company like ByteDance, which relies heavily on AI for content recommendation, moderation, and user analytics, this could force a rapid migration to Chinese AI hardware. See also: Alejandro Hernandez.

For Nvidia, the impact could be material. Losing a leading buyer like ByteDance erodes demand in what was one of its biggest growth markets. More broadly, the move could accelerate the adoption of home-grown AI processors across Chinese tech firms, reducing Nvidia’s influence in a critical global market.

This development may also prompt other countries to reconsider their dependence on a small set of foreign AI-chip suppliers. As China doubles down on self-sufficiency, we may see a rise in regional chip ecosystems globally — potentially fragmenting AI hardware standards worldwide. See also: Alejandro Garza.