Samsung AI chip profits reached a new global high in the second quarter of 2026, with preliminary earnings guidance placing operating profit at 89.4 trillion South Korean won, yet the result left markets unmoved as investors searched for signs the AI trade still has room to run.
The guidance, issued on 7 July 2026 and based on K-IFRS, put the operating profit range at 89.3 trillion to 89.5 trillion won and revenue between 170 trillion and 172 trillion won. The Korea Herald, citing a regulatory filing, calculated the year-on-year profit surge at 1,810.3 percent, with revenue up 129.3 percent to 171 trillion won. The report converts 89.4 trillion won to $58.7 billion; the Korea Herald put the same figure at approximately $61 billion, with the gap reflecting different exchange rates used at the time of publication.
The quarterly sequence has been relentless. Samsung posted operating profit of 20 trillion won in Q4 2025, then 57.2 trillion won in Q1 2026, before the Q2 figure delivered a further 56 percent jump from the prior quarter, according to the Chosun Ilbo. The same report notes that Samsung’s Q2 result exceeded Nvidia’s record quarterly operating profit of $53.5 billion (approximately 81.9 trillion won), which Nvidia posted in Q1 of its fiscal year 2027 covering February to April 2026, making Samsung’s result the largest quarterly operating profit among global private companies by that measure.
Samsung has published preliminary quarterly results since 2009, a practice designed to give investors early financial visibility, Aju Press reports.
What the Market Is Actually Asking
The scale of the numbers has not translated into conviction. Charu Chanana, chief investment strategist at Saxo, captured the shift in a Tuesday note: ‘The memory cycle is still strong, but the market is starting to ask whether the easy part of the trade is already behind us.’
Investors are no longer satisfied with strong sales alone, Chanana added. They want confident guidance, durable pricing power, and evidence that AI demand is not approaching a peak. Her concern was structural: ‘The question is no longer whether memory demand is strong. It is whether today’s shortage could eventually become tomorrow’s overcapacity problem if supply comes back too aggressively.’
The dynamic is familiar to anyone who has followed Nvidia through the same cycle. When Nvidia reported its fiscal third-quarter results in November 2025, beating expectations on both sales and earnings, CNBC reported that chief executive Jensen Huang opened by rejecting the premise of an AI bubble on the analyst call. In comments to employees after that quarter, Huang put the bind plainly: ‘If we delivered a bad quarter, it is evidence there’s an AI bubble. If we delivered a great quarter, we are fueling the AI bubble.’
Samsung AI Chip Profits in Context: A Trade That Priced in Perfection
James Thorne, chief market strategist at Wellington-Altus, framed the broader problem on X on Monday. ‘That is what happens when a bottleneck trade gets crowded: fundamentals stay strong, but earnings stop impressing because perfection was already priced in,’ he wrote.
Thorne argued the next leg of the AI buildout could shift rewards away from chipmakers and towards power, grid capacity, cooling, and physical infrastructure. His conclusion was blunt: ‘Artificial intelligence is not over. But easy trade in it is.’
The memory market underpins both the euphoria and the anxiety. Samsung AI chip profits have been driven by specialised memory chips used to train and run AI models, and analysts expect demand to continue outpacing supply for now. But Chanana’s overcapacity warning reflects a well-worn pattern in the memory industry, where shortage cycles have historically attracted aggressive capacity additions that eventually flip the market.
The immediate question is whether Samsung’s formal second-quarter results, expected later this month, deliver the detailed guidance investors are looking for. Without it, a profit figure that once would have seemed unimaginable may keep doing exactly what it did this week: failing to impress.