Nvidia Launches Vera Rubin Platform and Expands Taiwan Investment

The Nvidia Corp. has launched the Vera Rubin platform, featuring custom Vera CPUs designed to accelerate agentic AI workloads. In late May 2026, the company began delivering initial hardware racks to major AI firms, including OpenAI, Anthropic, and SpaceXAI. CEO Jensen Huang reported that the platform targets a $200 billion addressable market, with standalone CPU revenue expected to reach $20 billion this year. This expansion follows a record fiscal 2027 first-quarter revenue of $81.6 billion.

To secure its supply chain, Nvidia is dramatically increasing its footprint in Taiwan. Huang announced that annual spending in the region has grown to $150 billion, a tenfold increase over five years. The company is constructing a first-of-its-kind overseas headquarters, Nvidia Constellation, at the Beitou-Shilin Technology Park in Taipei, which is expected to be operational by 2030 and employ 4,000 people.

Despite this growth, the company faces significant geopolitical hurdles. Huang confirmed that Nvidia has largely lost its Chinese AI chip market to Huawei due to US export restrictions. While the US government licensed H200 chips for some Chinese firms, deliveries have stalled because Chinese officials have withheld approval. Additionally, the company is managing legal fallout from a US Justice Department investigation into the illegal export of $2.5 billion in AI technology to China via partner employees. In a move to maintain academic and professional ties, Huang recently agreed to join the advisory board of the Tsinghua University School of Economics and Management in Beijing.