The demand for artificial intelligence is driving the growth of the high bandwidth memory (HBM) market.
The demand for artificial intelligence is driving growth in the high bandwidth memory (HBM) market, and Samsung Electronics' AI business profitability prospects are attracting attention, but the gap between its valuation and that of Nvidia remains a focus of market discussion.

Taipei — The artificial intelligence (AI) boom is creating major opportunities across the global semiconductor industry, but stronger revenue and earnings do not necessarily translate into an equivalent increase in market valuation. A Reuters Breakingviews analysis said Samsung Electronics’ earnings outlook has improved sharply on strong demand for memory chips. Even so, investors continue to assign Samsung a much lower forward price-to-earnings ratio than Nvidia, reflecting differing views of the companies’ business models, capital spending and the durability of their future growth.

Samsung Electronics on Oct. 8 forecast third-quarter operating profit of about US$80 billion. Analysts expect the company’s full-year revenue to reach US$282 billion, compared with an estimated US$268 billion for Nvidia. These figures, however, reflect Samsung’s quarterly earnings guidance and market forecasts for full-year revenue, not the companies’ final reported results for the year.

AI memory demand lifts Samsung’s earnings outlook

Samsung is one of the world’s leading memory chipmakers. The expansion of AI data centers is driving demand for dynamic random-access memory (DRAM), while high-bandwidth memory (HBM) has become a key component in AI accelerators.

Breakingviews, citing Bernstein analysts, said average DRAM prices rose about 20% in the quarter ended September from the previous quarter. Analysts also expect prices to increase another 10% to 15% in the December quarter. Traditional DRAM is used in products including personal computers and smartphones, while HBM is widely used in high-performance chips that support AI computing.

Higher memory prices are helping improve Samsung’s earnings, but these products are also subject to industry cycles. Corporate profits can come under pressure when supply increases, demand weakens or prices decline. As a result, the market is assessing not only current AI demand but also how long the growth wave can last.

Why is Samsung valued below Nvidia?

According to Reuters Breakingviews, Samsung’s forward price-to-earnings ratio is about four times projected earnings for the next year, compared with roughly 17 times for Nvidia. The gap reflects the market’s differing assessment of the two companies, although the price-to-earnings ratio alone does not determine which is the more attractive investment.

Samsung’s businesses span memory, foundry services, smartphones, televisions and other electronic products. This diversification helps spread some risks, but it also means the company must contend with competition, operational demands and investment requirements across different markets.

By contrast, Nvidia is focused on AI computing chips and related platforms, while outsourcing chip manufacturing to foundry operators including Taiwan Semiconductor Manufacturing Co. Reuters Breakingviews said memory products are more exposed to price cycles, whereas analysts still see room for Nvidia’s earnings expectations to grow.

Analysts cited by the column forecast that Samsung’s operating profit could peak in 2029. Nvidia’s operating profit, meanwhile, is expected to continue rising over the next five years, although its growth rate may slow. These are forecasts, not confirmed outcomes.

Heavy capital spending and pressure from shareholders

Another challenge for Samsung is its substantial capital spending. Reuters Breakingviews said Samsung’s capital expenditure in 2027 is expected to exceed US$64 billion, more than five times Nvidia’s projected spending.

Samsung needs to continue investing in factories in South Korea and the United States while also supporting other businesses, including smartphones and televisions. Such spending helps expand capacity and maintain competitiveness, but it also affects cash flow, investment payback periods and shareholder returns.

Samsung must also balance employee compensation with shareholder interests. The column said Samsung reached a tentative agreement with its labor union in May to link bonuses to operating profit, averting a possible strike involving about 48,000 employees. Some minority shareholders, however, have questioned the compensation arrangements and taken legal action.

Regarding its share price and shareholder returns, the column said Samsung shares have still risen substantially this year but have fallen by more than a quarter from their June peak. Some investors have also called on the company to increase returns to shareholders.

South Korea expands semiconductor investment plans

The opportunities created by AI are affecting not only companies but also South Korea’s industrial policy. Reuters Breakingviews said the South Korean government recently proposed a large-scale semiconductor investment strategy involving Samsung and rival SK Hynix, with total investment exceeding US$576 billion.

The move shows that South Korea is seeking to capitalize on AI-driven chip demand and strengthen its domestic semiconductor industry. But larger investments do not guarantee a corresponding increase in future profits. The actual results will depend on market demand, technological competition, capacity utilization and the recovery of invested capital.

Overall, Samsung’s earnings outlook highlights the opportunities created by demand for AI memory, while its valuation and capital-spending challenges underscore the differences among business models in the semiconductor industry. For readers in Taiwan, the comparison between Samsung and Nvidia also helps illustrate the distinct roles of memory manufacturing, AI chip design and TSMC’s foundry operations in the global AI supply chain.