Summary of Semiconductor Industry Conference Call Industry Overview - The U.S. semiconductor industry, particularly in AI computing, has seen GPU and ASIC production peak but plateau since January 2025, with a subsequent decline in output. The slowdown of Moore's Law and the deceleration of NVIDIA's GB200 have contributed to a contraction in computing demand, compounded by the slower-than-expected commercialization of popular applications [1][3]. Key Points and Arguments - **Market Performance**: In Q1 2025, both AI applications and AI computing companies experienced a decline, with applications suffering a larger drop. Leading companies like NVIDIA, TSMC, and Broadcom showed strong rebounds, supported by macroeconomic stability and improved corporate performance [1][5]. - **Macroeconomic Factors**: Recent weeks have seen stock price changes in the AI computing sector influenced by macro variables, earnings guidance from major firms, and potential orders in the industry [1][7]. - **TSMC Expansion**: TSMC plans to add nine new factories this year, with 2nm production expected to start in the second half of 2025 and 3nm already at full capacity, indicating optimism about future AI computing demand [1][9]. - **Intel's Developments**: Intel's new CEO is pushing for advancements in wafer foundry services, with the 18A process set to begin production. Breakthroughs in advanced processes could enhance Intel's future performance [1][11]. Additional Important Insights - **Impact of Deepseek**: The emergence of Deepseek has led to a contraction in computing demand, alongside the slow commercialization of popular applications, which has collectively suppressed demand for computing power [2][5]. - **Capital Expenditure Trends**: AI computing demand remains stable, with major firms expected to see capital expenditure growth of over 30% year-on-year, although this growth rate is slowing [4][16]. - **Future Demand Predictions**: The AI computing sector is expected to see a potential uptick in training demand in the next 2-3 quarters, despite current weak expectations. Companies like Tencent and Alibaba are focusing more on inference demand through algorithm optimization [20]. - **Upcoming Events**: Key upcoming events include NVIDIA's earnings report and major tech conferences, which are expected to provide insights into industry developments and potential shifts in demand [21]. Conclusion - The semiconductor industry, particularly in AI computing, is navigating a complex landscape influenced by macroeconomic factors, corporate performance, and technological advancements. The outlook remains cautiously optimistic, with potential growth driven by leading companies and upcoming technological releases.