In June, Samsung Electronics and domestic rival SK Hynix laid down a huge marker for the AI era: a combined 800 trillion won ($599 billion) investment plan for a new chipmaking hub in southwest South Korea. The wager was straightforward — that data centres building out artificial-intelligence capacity will keep needing ever more memory.¹

By the July-to-September quarter, Samsung said that wager was paying off. The company reported operating profit of 107.4 trillion won ($80.5 billion), a record driven by demand for memory chips used in AI systems. Revenue rose to an all-time high of 195.3 trillion won ($146 billion), its fourth consecutive quarter of record profit and sales.²

The surge is not Samsung’s alone. SK Hynix reported record third-quarter profit a day earlier, underlining the extraordinary leverage held by the two South Korean groups, which together control about two-thirds of the global memory market.¹

Samsung’s message is that the boom still has room to run. It expects demand for AI-related chips to accelerate through 2027 and outstrip supply, focusing on high-bandwidth memory, server DRAM and solid-state drives for AI servers. “Soaring demand for memory chips” is fueling the company’s latest record, according to the report.¹

But the same concentration that makes the upswing lucrative could sharpen the downside. Analysts have questioned whether the scale of planned spending will pay off, while South Korea’s central bank has warned that heavier dependence on chips could magnify the blow from any AI-led downturn.¹