Nvidia's recent performance has alleviated concerns about a potential peak in the artificial intelligence (AI) industry. As fears of overinvestment and a bubble in AI grow, Nvidia, a leading company in the AI era, has once again demonstrated its growth potential.
On August 26, Nvidia announced that its revenue for the second quarter of the 2027 fiscal year reached $96.22 billion, significantly exceeding market expectations of $92 billion. The company also reported earnings per share of $2.22, surpassing forecasts. Additionally, Nvidia raised its revenue outlook for the third quarter to $108 billion. Notably, CEO Jensen Huang expressed strong confidence in the sustainability of AI demand, emphasizing that the establishment of AI infrastructure is at a critical turning point. He projected a 70% revenue growth by the 2028 fiscal year, signaling that the AI investment cycle will continue for a considerable time.
Concerns about a peak in AI investment have been persistent in the market. Questions have arisen about whether the astronomical investments in AI data centers will translate into actual profits. Criticism has also been directed at big tech companies for excessive AI investments, leading to speculation that semiconductor demand may have peaked and is set to decline. However, Nvidia's latest results clearly indicate that there are no signs of a downturn in AI infrastructure investment as of now.
In fact, the bottleneck lies not in demand but in supply. Nvidia noted that while strong demand continues, rising memory prices could impact profitability. This suggests that the next battleground in the AI semiconductor market is shifting from GPUs to memory. The importance of AI-specific memory, such as high-bandwidth memory (HBM) and high-performance DRAM, is expected to grow significantly.
For South Korean companies collaborating with Nvidia, this presents a substantial opportunity. The demand for memory, packaging, substrates, and power and cooling related to AI accelerators is likely to increase. In particular, this could serve as a crucial momentum for Samsung Electronics and SK Hynix to sustain the supercycle of AI semiconductors. Following Nvidia's earnings announcement, both Samsung and SK Hynix saw their stock prices rise in the domestic market, reflecting investor optimism.
However, Nvidia's strong performance does not completely dispel concerns about an AI bubble. There remains a need to verify whether the massive capital expenditures by AI companies will translate into actual profits. Even with good results, if market expectations are higher, stock prices may still fluctuate. There have been instances in the past where Nvidia's stock fell despite exceeding market expectations. It is essential to distinguish between the direction of the industry and corporate performance. While some bubble exists in the AI sector, the notion that demand for AI semiconductors has peaked is a separate issue. Nvidia's latest results indicate that such concerns have not yet materialized.
Leading in the AI semiconductor competition is crucial for national industrial competitiveness. Samsung Electronics and SK Hynix must enhance their research and development and capital investment to improve their next-generation memory competitiveness, including HBM. Issues related to power, water, and talent in the semiconductor infrastructure should not be left solely to companies. Now is the time to focus on securing technology and production capacity ahead of competitors while managing risks calmly, rather than delaying investments due to fears of an AI peak.
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.