Semiconductor Industry Shaken by AI Peak Concerns... Long-Term Outlook Remains Positive

Bubble Controversy Grows Over ASML and TSMC Earnings
Samsung Electronics and SK Hynix Also Linked to Stock Declines
Deloitte: "Need for AI Semiconductors Remains Unchanged"

As the era of artificial intelligence (AI) fully unfolds, cautious views have emerged in the semiconductor industry, which had been receiving positive news. After the earnings announcements from the Netherlands' ASML and Taiwan's TSMC, concerns arose that growth might fall short of market expectations. However, from a long-term perspective, many forecasts suggest that AI demand will be a key factor driving semiconductor growth.

On the 22nd, the semiconductor industry is focusing on the upcoming earnings announcement of Nvidia for the first quarter of fiscal year 2025 (February to April 2024). Amid debates about an AI bubble, the results from Nvidia, a leading U.S. semiconductor company and AI semiconductor market leader, will provide insights into the semiconductor market trends, outlook, and related latest figures.

Inside and outside the semiconductor industry, concerns have been raised that the semiconductor market, which recently showed remarkable growth due to AI effects, may have peaked. Reflecting this, Nvidia's stock price closed at $762 per share on the 19th (local time), a 10% drop from the previous day, and a 24.67% decline compared to January 25th ($950.02), all within less than a month.

Earlier, ASML, which monopolizes the supply of extreme ultraviolet (EUV) lithography equipment essential for advanced semiconductor processes, fueled market concerns by revealing that its new orders for lithography equipment in the first quarter amounted to 3.6 billion euros, only 33.3% of market estimates. TSMC, which manufactures Nvidia's AI semiconductors on a contract basis, also drew a line under optimism by revising downward its outlook for the non-memory and foundry markets during its first-quarter conference call last week.

Consequently, the impact has spilled over to Samsung Electronics and SK Hynix, which produce high-bandwidth memory (HBM) products used in AI semiconductors. There are forecasts that HBM prices could be adjusted downward in the second half of the year due to a slowdown in the market. Amid these concerns, the stock prices of Samsung Electronics and SK Hynix have continued to decline since last weekend.

The semiconductor industry expects that while market growth driven by AI effects may be lower than anticipated, the long-term impact will be significantly positive. Wei Zhejia, CEO of TSMC, stated in the recent conference call that demand for general servers, smartphones, and PCs remains sluggish, but he expressed a positive outlook regarding AI demand.

CEO Wei emphasized that "demand for AI-related data centers is very strong" and noted that "the shift of budgets from traditional servers to AI servers is advantageous for TSMC." He also forecasted that "the company's growth will be very healthy." TSMC expects that sales contributions from server AI processors, such as graphics processing units (GPUs), will more than double this year to account for just over 10% of total sales, and exceed 20% by 2028.

The market also holds a positive view regarding the long-term outlook. Deloitte Group recently projected in an insight report that the market size for generative AI-specific semiconductors will exceed $50 billion this year and could grow to between $110 billion and $400 billion by 2027. While conservative forecasts, including bubble concerns, could reduce market revenue, the market is still expected to be larger than current levels.

Deloitte Group stated, "Whether the market size is $100 billion or $400 billion, companies will still need AI semiconductors, especially generative AI semiconductors," adding, "The corporate recognition that securing a stable and reliable supply chain is essential to safeguarding innovation, economic success, and national security will not change."