Nvidia weighs expanding H200 output as China rushes to secure AI chips
Nvidia is considering ramping up production of its H200 artificial intelligence chips after seeing a surge in orders from Chinese technology companies. The U.S. chipmaker informed clients that requests have exceeded its current supply, following Washington’s decision to permit exports of the processors to China under a 25% tariff.
Major Chinese firms, including Alibaba, ByteDance, and Tencent, expressed strong interest in placing large-scale orders. The H200 chip delivers roughly six times the performance of Nvidia’s previously authorized H20, making it a highly sought-after component for powering advanced language models and data processing systems.
Beijing considers terms for approval
Chinese regulators are evaluating whether to allow imports of the H200 amid growing local demand. In meetings held on December 10 and 11, officials reportedly asked major tech companies to present their forecasts for AI computing needs. Authorities are also weighing a condition requiring companies to purchase a proportional quantity of domestic chips from suppliers such as Huawei and Cambricon alongside each H200 order.
The proposal highlights China’s effort to strengthen its semiconductor self-sufficiency while keeping pace with the global race for AI computing capacity. According to analysts, the H200 currently offers up to triple the performance of the most powerful chips produced locally.
Supply limits challenge Nvidia’s plans
Any production increase for the H200 could face delays due to supply constraints. Nvidia is prioritizing output for its next-generation Blackwell and Rubin architectures, leaving limited capacity for the older H200 line. Manufactured by TSMC using 4-nanometer technology, the chip has been in mass production since last year but remains in tight supply.
Industry experts say Nvidia must balance competing demands as global cloud providers and enterprises fight for AI chip allocations. The company is also contending with strong competition for manufacturing capacity at TSMC, shared with tech giants such as Google.
-
07:15
-
07:00
-
17:30
-
17:00
-
16:30
-
16:15
-
16:00
-
15:30
-
15:00
-
14:45
-
14:43
-
14:30
-
14:15
-
13:50
-
13:30
-
13:00
-
12:45
-
12:30
-
12:15
-
12:00
-
11:45
-
11:30
-
11:15
-
11:00
-
10:45
-
10:30
-
10:15
-
10:00
-
09:45
-
09:30
-
09:15
-
09:00
-
08:45
-
08:30
-
08:15
-
08:00
