Taiwan Semiconductor Manufacturing Corp. (TSMC) executives recently spoke out about the world’s growing hype over artificial intelligence (AI) and its impact on chip makers’ stock prices. While acknowledging the strong demand for AI processors, particularly in the data center sector, TSMC expressed caution regarding the sustainability of this demand. Furthermore, they made it clear that AI alone would not be sufficient to counterbalance the sluggish global demand for semiconductors.
The Short-Term Frenzy
TSMC Chairman Mark Liu emphasized during an earnings call with analysts that the current AI demand should not be extrapolated to the long term. He pointed out that predicting how this sudden demand will develop or plateau next year is challenging. In other words, while there is indeed a short-term frenzy surrounding AI demand, its future trajectory remains uncertain.
The Numbers Game
While AI-related processors currently account for 6% of TSMC’s revenue, the company recognizes that this alone cannot alleviate the overall macroeconomic impact of the semiconductor industry’s downturn. TSMC forecasts a 10% decline in its full-year revenue. However, they do anticipate a potential 50% compounded annual growth rate for AI processor demand over the next five years.
Skilled Workers and Production Delays
Aside from addressing AI demand, TSMC also discussed challenges related to its new manufacturing plant in Arizona. The production start date has been pushed back to 2025 due to difficulties in finding skilled workers for semiconductor equipment installation. As a solution, the company plans to send experienced technicians from Taiwan to provide training for local workers for a limited period.
In summary, TSMC’s executives urge caution when interpreting the current hype around AI demand. While acknowledging its short-term importance and growth potential, they emphasize the need for realistic expectations and a nuanced understanding of the overall semiconductor industry trends.
Chip Stocks Fall as Investors Question AI’s Role in Revenue Growth
The recent projection of a 50% compound annual growth rate (CAGR) for AI revenue has sparked uncertainty among investors. Needham & Co. analyst Charles Shi stated that management expects AI revenue to reach a low-teens percentage of total revenue by 2028; however, no additional information regarding these assumptions was provided.
Consequently, several semiconductor stocks experienced a decline, including high-performer Nvidia Corp. The company, known for its graphics processors (GPUs) utilized in compute-intensive AI applications, has seen an increased demand since the launch of ChatGPT. Despite forecasting a substantial boost in AI demand for its fiscal second quarter, with projected revenue of approximately $11 billion, Nvidia’s shares dropped by 2.42% on Thursday.
Other companies expected to benefit from chip sales for data centers, such as Intel Corp., Advanced Micro Devices Inc., and Broadcom Inc., also experienced a decline in their stock prices. The Philadelphia Semiconductor Index followed suit, dropping by 2.61%.
Furthermore, while TSMC executives acknowledge the growing demand for generative AI applications like ChatGPT, they did express concerns about the sustainability of current trends and the overall global demand for semiconductors. Investor focus shifted towards these concerns rather than the expected capital spending increase.
Evercore analyst C.J. Muse commented on the company’s outlook, stating that it has deteriorated compared to three months ago. This is primarily due to reduced optimism surrounding macro trends, a weaker-than-expected recovery in China, and ongoing customer inventory digestion.
The recent developments suggest that the chip downturn may not be over, challenging the belief that AI will be the ultimate savior in driving revenue growth for the semiconductor industry.
