Markets Wait for Impact of US Deals With Nvidia, AMD

Key Concepts:

• Chinese Market Access
• Global AI Ecosystem
• Dominant AI Standard
• Data as a Strategic Advantage
• US-China Tech Relationship
• Innovation Driven by Necessity
• Revenue Share (Quid Pro Quo)
• High Bandwidth Memory (HBM)
• Telematics
• Installed Capacity (Megawatt/Gigawatt)
• Policy Fluidity

China as a Strategic Market:

• China is considered a strategic market due to its scale and ecosystem.
• While the US is a significant source of technological innovation, China excels at large-scale implementation.
• Having a global ecosystem is a crucial long-term strategic advantage.
• More data leads to better AI models, emphasizing the importance of global data access.

The Quest for a Dominant AI Standard:

• The discussion revolves around who will establish the dominant AI standard globally.
• The ideal scenario is a unified global ecosystem, but the possibility of "forks" (separate, incompatible systems) exists.
• The speaker references statements from Jensen (likely Jensen Huang, CEO of NVIDIA) and Lisa Su (CEO of AMD) regarding this competition.

Impact of Export Restrictions and Innovation:

• The impact of restrictions on exporting technology to China is discussed.
• Necessity drives innovation: If exports are completely cut off, China will likely develop its own independent ecosystem.
• The current situation is viewed as a bridge, potentially mitigating the drastic actions seen earlier in the year.

China's Advantages:

• China possesses advantages, including a newly developed power infrastructure and vast amounts of data.
• They are rapidly advancing open-source AI models.

Importance of Global Standards and Data Access:

• Maintaining a global standard is crucial for US companies.
• Examples like the iPhone and Tesla are cited, where data resides in China but lessons are learned from adoption.
• While existing AI models have been trained on substantial data, new data is constantly generated through application usage.
• Access to this new data is essential for continued AI development.

Revenue Share and Market Access:

• The Treasury Secretary's suggestion of a "pay to play" or "quid pro quo" model (revenue share for market access) is discussed.
• The question is raised whether this model could expand to areas like high bandwidth memory (HBM) or telematics.
• If such a model were implemented, it could potentially lead to an upgrade in the outlook for installed capacity (measured in megawatts or gigawatts).

Policy Fluidity and Uncertainty:

• It's too early to determine the impact of these potential changes.
• Policy has been fluid, with frequent changes.
• While some of the most drastic assumptions from earlier in the year are being discounted, it's premature to make aggressive assumptions.
• The current situation is viewed as an improvement from the worst-case scenario.

Conclusion:

The discussion centers on the strategic importance of the Chinese market, the competition for establishing a dominant AI standard, and the impact of US export policies. Access to data and the potential for revenue-sharing agreements are key factors influencing the future landscape. However, significant uncertainty remains due to policy fluidity and the evolving nature of the US-China tech relationship.