Trump Discussed Nvidia Chips With Xi Jinping | Bloomberg Tech 5/15/2026

Key Concepts

• Semiconductor Supply Chain: The global reliance on Taiwan (TSMC) versus the push for domestic U.S. manufacturing (CHIPS Act).
• AI Infrastructure Cycle: The massive capital expenditure (capex) by hyperscalers (Alphabet, Amazon) to build data centers and custom silicon.
• Humanoid Robotics: The shift from lab-based prototypes to autonomous, 24/7 commercial operations (Figure AI).
• Vertical Integration: The strategy of owning the entire stack—from hardware/actuators to neural network training—to achieve scale.
• Design Stack Disruption: The debate over whether AI tools (like Figma’s AI features) replace or augment human design workflows.

---

1. Semiconductor Industry & Geopolitics

• Market Impact: The Philadelphia Semiconductor Index (SOX) saw a significant decline (nearly 4%) as markets digested the meeting between President Trump and President Xi.
• Nvidia H200: President Trump confirmed discussions regarding Nvidia’s H200 chips with President Xi. However, no purchase agreements were finalized, as China is prioritizing the development of its own domestic chip capabilities.
• Workforce Bottlenecks: Sheri Les (SEMI) highlighted that the U.S. faces a shortage of ~150,000 skilled workers. The CHIPS Act is funding a $200 million initiative via the National Science Foundation to build regional workforce nodes.
• Manufacturing Reality: While the U.S. is building new fabs (e.g., TSMC in Arizona), the industry requires a diverse talent pool, ranging from PhD researchers to fab floor operators and technicians.

2. AI Infrastructure & Hyperscaler Capex

• Investment Trends: Eric Sheridan (Goldman Sachs) noted that we remain in an infrastructure cycle. While capex growth is moderating, the combined revenue backlog for Alphabet and Amazon’s cloud divisions exceeds $900 billion.
• Custom Silicon: Alphabet (TPUs) and Amazon are increasingly using in-house chips. These offer competitive price-to-performance ratios compared to standard GPUs, allowing hyperscalers to capture more margin and drive cloud ecosystem adoption.
• Interdependence: The AI landscape is becoming highly interdependent, with only a handful of companies possessing the capital and scale to build the necessary infrastructure.

3. Corporate Developments & IPOs

• Cerebras IPO: Following a massive debut (up to 89% at one point), the stock saw profit-taking. Early venture backers (Benchmark, Eclipse, Foundation Capital) are poised to make billions.
• Figma’s Growth: Despite fears of AI disruption, Figma reported 46% revenue growth. CEO Dylan Field emphasized that AI is being used to "mold" model outputs (e.g., Figma Weave), turning design into a more collaborative, workflow-based process.
• Figure AI: CEO Brett Adcock confirmed that their humanoid robots have operated autonomously for over 50 hours on a logistics line, sorting 60,000+ packages. The robots use an in-house neural network (Helix 2) and are manufactured at "Bot Q" at a rate of 60–70 units per week.

4. OpenAI: Legal & Operational Challenges

• Apple Relationship: Reports indicate the partnership is fraying, with OpenAI weighing legal action, claiming Apple has limited the integration of its technology, thereby failing to deliver expected subscriber growth.
• Musk vs. OpenAI Trial: Closing arguments concluded in the lawsuit brought by Elon Musk. Musk is seeking $134 billion in damages, the removal of Sam Altman and Greg Brockman, and the conversion of OpenAI back to a nonprofit. The jury’s verdict will be advisory to the judge.
• Leadership Dynamics: CFO Sarah Frier described her relationship with Sam Altman as a "ying and yang" partnership, focused on managing the massive compute demand and ensuring financial optionality for the company.

---

Synthesis/Conclusion

The tech landscape is currently defined by a massive "infrastructure race." While geopolitical tensions regarding Taiwan and China create market volatility, the underlying trend is a relentless push toward vertical integration. Companies like Amazon and Figure AI are moving to own their entire technology stacks—from custom silicon to robotics hardware—to mitigate supply chain risks and capture margins. Meanwhile, the "AI disruption" narrative is being challenged by companies like Figma, which are successfully monetizing AI as a productivity multiplier rather than a replacement for human expertise. The week concludes with significant legal and structural uncertainty surrounding OpenAI, highlighting the high-stakes nature of the current AI arms race.