Samsung Electronics has put an extraordinary number on the artificial-intelligence boom: about $80 billion in operating profit for a single quarter.
In preliminary guidance released Thursday, October 8, Samsung estimated third-quarter revenue of 195 trillion won and operating profit between 107.3 trillion and 107.5 trillion won. At the midpoint, that is 107.4 trillion won, or roughly $80.2 billion at the exchange rate used by Reuters.
The figure is not yet a final audited result. But if it holds when Samsung releases its detailed earnings on October 29, Reuters reported it would be the largest quarterly operating profit ever posted by a technology company—and nearly nine times Samsung’s operating profit from the same period a year earlier.
The number is bigger than the headline
Samsung’s official earnings guidance gave a narrow operating-profit range of 107.3 trillion to 107.5 trillion won. The midpoint was slightly above the 106.1 trillion won average estimate compiled by LSEG, according to Reuters.
Yonhap News Agency calculated that the projected operating profit represents a 782.5% increase from 12.17 trillion won a year earlier. Revenue of 195 trillion won would be up 126.6% year over year.
That matters because operating profit is the money Samsung expects to earn from its core businesses before interest and taxes. It is not the same as net income, and Samsung has not yet disclosed the division-by-division details that will show exactly how much came from semiconductors, smartphones, displays and appliances.
AI is turning memory into strategic infrastructure
The central force behind the forecast is the scramble for memory used in AI data centers. Advanced systems require huge quantities of high-bandwidth memory, or HBM, alongside conventional DRAM and storage chips. As hyperscale cloud companies and AI developers race to deploy more computing power, demand has risen faster than manufacturers can add qualified supply.
That imbalance has lifted prices and improved margins across the memory industry. Samsung, SK Hynix and Micron are the dominant suppliers, and Reuters reported that Samsung and Micron expect the tight market to persist into 2028.
For Samsung, the quarter would mark a dramatic payoff from efforts to regain momentum in HBM. The company has invested heavily in advanced packaging and next-generation memory while continuing to operate one of the world’s broadest semiconductor portfolios. Its scale means it can benefit not only from premium AI components but also from stronger pricing across mainstream memory products.
Why investors will wait for October 29
Preliminary guidance answers the biggest question—how profitable the quarter was—but leaves several important ones open.
- How much came from chips? Samsung will disclose semiconductor revenue and profit only with the full results.
- How durable are today’s prices? Memory remains a cyclical business, and a sudden slowdown in AI infrastructure spending could change the balance quickly.
- Can Samsung sustain HBM progress? Qualification timelines, manufacturing yields and competition will determine whether it can convert industry demand into lasting share gains.
- What did the boom cost? Investors will watch capital spending and free cash flow as Samsung expands production capacity.
The full report will also reveal how Samsung’s mobile and consumer-electronics operations performed while the semiconductor business captured the spotlight.
A milestone—and a warning about the scale of the AI race
The projected result is a milestone for Samsung, but it also says something broader about the economics of AI. The most visible products may be chatbots and software services, yet a vast share of the money is flowing into the physical layer beneath them: memory, processors, networking gear, power systems and data centers.
An $80 billion quarterly operating profit would show just how valuable a supply bottleneck can become when nearly every major technology company is chasing the same scarce components at once. It would also raise the stakes for customers, rivals and regulators watching the concentration of the global chip supply chain.
For now, Samsung’s number remains a forecast rather than a final result. The October 29 filing will determine whether the company has officially set a new benchmark—and how much of that benchmark rests on an AI memory market still running hotter than supply can follow.
