In May 2026, NVIDIA officially established the 800V high-voltage DC architecture as the core power solution for next-generation AI factories. A recent Bank of America report noted that as NVIDIA’s GPU platforms evolve, data center rack power consumption will surge from the traditional server range of 10–15 kilowatts to over 1.5 megawatts during the Feynman platform era in 2029–2030—an increase of nearly 100-fold. The total addressable market (TAM) for analog semiconductors in AI data centers is projected to expand from $7.9 billion in 2025 to approximately $27 billion by 2030, representing a five-year compound annual growth rate (CAGR) of 28%. Analog chipmakers stand to benefit most directly, with wide-bandgap semiconductors—specifically silicon carbide (SiC) and gallium nitride (GaN) power devices—accelerating their migration from the automotive industry into AI data centers. Among Hong Kong- and U.S.-listed power semiconductor companies, which do you favor?