Nvidia has stopped producing its H200 artificial intelligence chips for the Chinese market. The decision comes as export restrictions between the United States and China remain uncertain. The company is now redirecting manufacturing capacity toward its next generation Vera Rubin chips.
The move affects production at Taiwan Semiconductor Manufacturing Company, which builds Nvidia’s advanced processors. Manufacturing lines that were previously allocated to H200 chips are now being shifted toward the Vera Rubin platform.
Around 250000 H200 units have already been produced. Those chips could cover limited demand if sales eventually resume in China. However analysts say meaningful sales in the Chinese market are unlikely in the near term.
Nvidia stock remains closely watched as investors evaluate the potential impact on AI chip demand and global semiconductor supply chains.
Nvidia H200 China production halt amid export restrictions
Nvidia’s decision to halt H200 production for China reflects the ongoing uncertainty around export controls. The United States has imposed strict limits on advanced AI chip exports to Chinese companies.
Even though some licenses have recently been granted, the regulatory situation remains unclear. Because of this uncertainty Nvidia is not expecting significant H200 sales in China anytime soon.
China is also pushing hard to develop its own domestic semiconductor industry. Chinese companies are investing heavily in homegrown AI chips to reduce reliance on US technology.
Talks between the United States and China about technology trade policies could take place later in March. Until clearer rules emerge Nvidia appears to be limiting exposure to the Chinese market.
Nvidia shifts TSMC capacity to Vera Rubin chips
The halt in H200 production is leading to a shift in manufacturing resources. Nvidia has redirected production capacity at TSMC toward its upcoming Vera Rubin chip architecture.
This transition is expected to last around 3 months. During that period the company will focus on ramping up production for next generation AI processors.
The shift means there will likely be no immediate revenue boost from China. Instead Nvidia is prioritizing future chip platforms that could power the next wave of artificial intelligence computing.
With roughly 250000 H200 units already produced Nvidia still holds some inventory. Those chips could meet limited demand if export approvals expand later this year.
For now the company’s strategy signals a clear shift toward next generation AI hardware while waiting for clarity on US China semiconductor trade rules.