On the right is Banghua Zhu, Chief Research Scientist at NVIDIA
AsianFin -- Silicon Valley’s top tech firms are intensifying a high-stakes race to secure Chinese AI talent, with Nvidia and Meta leading the charge through aggressive hiring and strategic acquisitions aimed at cementing their lead in the global artificial intelligence arms race.
On Sunday, Nvidia CEO Jensen Huang brought two rising Chinese stars into his orbit: Banghua Zhu, a Tsinghua University alumnus and assistant professor at the University of Washington, and Jiantao Jiao, a UC Berkeley professor and fellow Tsinghua graduate.
Both co-founded cybersecurity-focused AI startup Nexusflow, which Nvidia appears to have quietly acquired, bringing Zhu on board as Chief Research Scientist for its Star Nemotron team and appointing Zhang Jian, another Tsinghua alumnus and Nexusflow CTO, as Director of Applied Research.
Meta, meanwhile, has lured away four senior researchers from OpenAI—Shuchao Bi, Jiahui Yu, Hongyu Ren, and Shengjia Zhao—who were instrumental in OpenAI’s multimodal and foundation model research, including the GPT-4o family. Their defections, first reported by The Information and seemingly confirmed by a since-deleted internal post, underscore the erosion of OpenAI’s once-dominant position in the AI talent hierarchy.
The hiring frenzy follows a broader shift in AI strategy by major tech firms. Meta recently invested \$14.3 billion for a 49% stake in Scale AI, onboarding founder Alexandr Wang to lead a new superintelligence research lab. Nvidia, for its part, has outpaced its own record acquisition spree, buying 26 companies or projects so far in 2025—most in the U.S. and Israel—and pumping over \$1 billion into AI investments this year.
Nvidia has also integrated several acquisitions into its ecosystem. LeptonAI, founded by former Alibaba VP Jia Yangqing, now powers Nvidia's DGX Cloud Lepton. CB Insights notes that Nvidia is leveraging GPU supremacy to drive AI infrastructure expansion, targeting future growth areas like quantum computing and 5G.
Bank of America estimates Nvidia holds 80%-85% of the AI data center chip market. Analyst Vivek Arya recently raised Nvidia's FY2027 EPS forecast to \$7.23, citing strong fundamentals. Barclays predicts Nvidia could hit a \$5 trillion market cap with its new Blackwell platform.
The surge in Chinese AI talent within top U.S. tech firms reflects a deeper shift. While the internet era saw Indian executives rise to dominance, the AI era is witnessing a parallel surge in Chinese leadership. Nvidia’s Huang, AMD’s Lisa Su, Broadcom’s Hock Tan, and Intel’s Pat Gelsinger all trace roots to Greater China.
Jensen Huang himself has praised China’s AI progress, noting that 50% of global AI researchers are Chinese. “You can't stop them,” he said, referencing 『DeepSeek』's impressive model performance. “Huawei is a world-class company, and Chinese researchers are top-tier.”
Meanwhile, the geopolitical backdrop is heating up. With a new administration in Washington, former President Donald Trump is pushing to dominate the AI race. Reuters reports the White House is preparing an aggressive AI infrastructure plan, easing grid access for data centers and allocating land for development. A formal U.S. AI strategy may launch as early as July 23.
The AI-driven energy demand is surging. Grid Strategies estimates U.S. electricity demand will grow fivefold by 2029, while Deloitte forecasts AI data centers could increase their energy usage thirtyfold by 2035.
Earlier this year, the White House unveiled the $100 billion "Stargate" project led by OpenAI, SoftBank, and Oracle. Trump’s stated goal: outpace China and make the U.S. the global hub of artificial intelligence.
Still, many Chinese experts see a need for cooperation. Tsinghua University’s Xiao Xi and Tang Xinhua warn that AI competition should not preclude dialogue and shared progress. “This is a marathon, not a sprint,” said Xiao.
Yet the current wave of AI hiring in Silicon Valley suggests that the sprint is on—and the frontrunners are fighting harder than ever to win.
