Core Insights - TSMC's Arizona plant has successfully produced its first batch of chips for major clients including Apple, NVIDIA, and AMD, marking a significant milestone in its operations [1][2] - The plant is currently utilizing N4 process technology and has already shipped 20,000 wafers, with NVIDIA's Blackwell AI GPU chips being a notable product [1][2] - TSMC's packaging capacity is a critical bottleneck in the AI chip supply chain, with plans to expand production from 75,000 to 115,000 units this year [1][2] Production and Capacity Expansion - TSMC's Arizona facility has produced 20,000 wafers, including those for NVIDIA's AI chips, which will undergo advanced packaging in Taiwan using CoWoS technology [2] - The company plans to expand its production capabilities to include 3nm and 2nm processes in future facilities [2] - TSMC is collaborating with Amkor to develop advanced packaging capabilities in the U.S., aiming to eventually complete packaging domestically [1][2] Market Demand and Client Engagement - The demand for AI chip packaging has led TSMC to increase its production capacity, attracting interest from other companies like UMC, which is collaborating with Qualcomm [2] - Major U.S. companies, including Apple, NVIDIA, AMD, and Qualcomm, are securing production capacity at TSMC's U.S. facilities due to geopolitical considerations and the need for backup production [4][5] - TSMC's new factories in the U.S. are reportedly fully booked, reflecting strong demand from clients [4][5] Investment and Future Plans - TSMC announced an additional investment of $100 billion in the U.S., bringing its total investment to $165 billion, which will support the establishment of six chip factories and two advanced packaging facilities [5] - The first factory in Arizona is set to begin mass production of advanced 4nm chips, with plans for subsequent factories to produce the latest 2nm chips by 2028 [5] - TSMC's chairman indicated that 30% of the advanced process capacity below 2nm is expected to be produced in Arizona to meet U.S. client demands [5]