Investing.com -- NVIDIA Corporation (NASDAQ: NVDA) announced Tuesday it is resuming manufacturing of AI chips for the Chinese market, according to CEO Jensen Huang, who also unveiled a $1 trillion order projection for the company's Blackwell and Vera Rubin chips through 2027 at the GTC 2026 conference this week. The stock declined 0.70% to $181.96 on Tuesday despite the bullish revenue forecast.

Since then, Nvidia has received licenses to export the H200 from the U.S. government and has taken orders, Huang said. This led Nvidia to begin restarting its manufacturing several weeks ago. "Our supply chain is getting fired up," Huang said.

The announcement marks a significant strategic shift for the world's most valuable public company, valued at approximately $4.5 trillion, as it navigates a complex regulatory environment surrounding AI chip exports to China. The decision comes after Nvidia faced a $5.5 billion financial hit from U.S. export restrictions and Chinese regulatory challenges over the past year.

Huang's $1 trillion revenue projection represents more than double the $500 billion forecast from a year earlier, underscoring the explosive growth in global AI demand. "In fact, we are going to be short. I am certain computing demand will be much higher than that," Huang said during the keynote address.

The timing of the China restart coincides with evolved U.S. export control frameworks. In December 2025, the Trump administration announced that Nvidia's H200 chip could be exported to approved customers in China if sales meet licensing conditions and the U.S. government receives 25% of the revenue.

Nvidia has faced mounting pressure on multiple fronts in China. Chinese regulators accused the company of violating anti-monopoly law in September 2025, related to conditions from its 2020 Mellanox acquisition. Additionally, China banned foreign AI chips from state-funded data centers in late 2025 while encouraging domestic chip development.

At GTC 2026, Nvidia also unveiled its next-generation Vera Rubin chips, designed to operate 3.5x faster than Blackwell on model-training tasks and 5x faster on inference tasks. The company announced its $20 billion acquisition of chipmaker Groq in December 2025 and introduced the Nvidia Groq 3 Language Processing Unit, scheduled to ship in the second half of 2026.

The shift to agentic AI systems that spawn multiple agents to accomplish tasks has dramatically increased inference workloads, driving demand for faster processing chips beyond traditional training-focused GPUs. Huang also announced partnerships with Hyundai, Nissan, and Chinese automakers BYD and Geely to build 18 million robotaxis annually, with Nvidia-powered self-driving Ubers expected to launch in 2027.

What to Watch

Investors should monitor several key developments:

• Regulatory clarity on compliance with U.S. export controls and the 25% revenue-sharing framework for China sales

• Production timeline and volume commitments for the restarted China chip manufacturing

• Revenue impact from China market re-entry on quarterly earnings, given the previous $5.5 billion hit from restrictions

• Customer adoption of Vera Rubin chips launching later this year, particularly the claimed 3.5x-5x performance improvements

• Groq 3 LPU shipments beginning in second half 2026, representing the first products from Nvidia's largest acquisition

• Robotaxi partnership progress with automakers toward the 18 million annual production target