AI Weekly: Nvidia's OpenAI deal wait and mega chip plants | REUTERS

Key Concepts:

• Nvidia-OpenAI Deal
• Circular Deals in AI Ecosystem
• Anthropic IPO
• Micron Technology AI Memory Chip Plant in Japan
• Chinese AI Firms Training Abroad
• Taiwan's AI-Driven Economic Growth

Nvidia's Open AI Deal and Investor Concerns

Nvidia's Chief Financial Officer (CFO) has stated that the company's proposed $100 billion investment in OpenAI, the parent company of ChatGPT, is not yet finalized. This potential partnership is significant as it would link two major players in the artificial intelligence (AI) race. However, the deal is also drawing scrutiny from investors due to concerns about "circular deals" within the AI ecosystem. These concerns arise because Nvidia is investing heavily in companies that are also its major customers. The proposed deal with OpenAI reportedly involves deploying at least 10 gigawatts (GW) of systems, which is an amount of power sufficient to supply over 8 million US homes.

Anthropic's Potential IPO and Funding Round

According to a report by the Financial Times, Anthropic, an AI startup supported by Google and Amazon, is reportedly preparing for an Initial Public Offering (IPO) as early as next year. The same report suggests that Anthropic is in negotiations for a private funding round that could value the company at over $300 billion. However, a spokesperson for Anthropic has not confirmed to Reuters whether or when the company plans to go public.

Micron Technology's Investment in Japan

Micron Technology is planning a significant investment of $9.6 billion in Japan to establish a new AI memory chip plant. This information was reported by Japan's Nikkei, citing sources familiar with the matter. The plant is expected to be located in Hiroshima, in western Japan, and will focus on producing advanced high bandwidth memory (HBM) chips. According to the report, Micron aims to commence construction in May of the following year and begin shipping these chips around 2028.

Chinese AI Firms Training Models Abroad

Top Chinese technology firms are reportedly training their AI models outside of China, according to the Financial Times. This strategy is being employed to gain access to Nvidia's advanced chips and to circumvent US measures designed to restrict their progress in cutting-edge technology. The Financial Times, citing sources, indicated that companies such as Alibaba and ByteDance are among those training their latest large language models (LLMs) in data centers located in Southeast Asia. Reuters has not yet been able to independently verify this report.

Taiwan's Economic Growth Fueled by AI Demand

Taiwan's economy, heavily reliant on technology, is projected to experience its fastest growth in 15 years, driven by the surge in demand for AI technology. This forecast comes from the country's statistics office. The office anticipates that the Gross Domestic Product (GDP) will expand by over 7.3% this year, a substantial increase from the approximately 4.5% growth predicted in August. However, the outlook for the following year may be impacted by US tariffs.

Synthesis/Conclusion

The AI landscape is characterized by significant strategic investments and evolving market dynamics. Nvidia's substantial proposed deal with OpenAI highlights the growing interdependence between chip manufacturers and AI developers, while also raising investor concerns about the structure of these partnerships. Meanwhile, emerging AI companies like Anthropic are exploring public market options, and established players like Micron are expanding their manufacturing capabilities in key regions like Japan to meet the escalating demand for AI-specific hardware. The global race for AI dominance is also evident in the strategies of Chinese firms seeking access to critical technology abroad, and in the economic benefits reaped by economies like Taiwan that are central to the AI supply chain. However, geopolitical factors, such as US tariffs, continue to present potential headwinds for future growth.