|
Getting your Trinity Audio player ready...
|
Samsung Electronics has told investors that the artificial intelligence build-out is now showing up in its profit line on a scale few companies have ever seen. In preliminary guidance released on Thursday, October 8, the South Korean group said it expects third-quarter operating profit of about 107.4 trillion won, roughly $80 billion, up from 12.17 trillion won in the same period last year. Revenue is estimated at around 195 trillion won, according to SammyGuru’s summary of the filing, compared with 86.1 trillion won a year earlier. The company has not released a division-by-division breakdown, and its full results are due later this month.
The headline figure needs a careful reading. It is operating profit, which is earnings from business activities before interest, taxes and other items, and it is preliminary and unaudited. Some outlets have described it as the highest quarterly profit of any technology company, but I could not find a primary-source comparison that confirms that claim, and measures such as net income and operating income are not interchangeable. What the reports do establish is that it is a record for Samsung, its fourth consecutive record quarter, and, according to TrendForce, the first time a South Korean company has earned more than 100 trillion won in operating profit in a single quarter.
The size of the jump is striking. The Next Web notes that the profit is nearly nine times the year-earlier figure and beat the LSEG SmartEstimate of 106.1 trillion won. It also tops Samsung’s own previous record from the second quarter, when operating profit was 89.49 trillion won on sales of 171.5 trillion won. One data provider pointed out that the quarter’s profit alone exceeds the company’s entire operating profit for 2025, which was 43.6 trillion won. Revenue, meanwhile, has more than doubled year on year, which suggests the increase comes from higher prices and volumes at the same time, not from cost-cutting.
The engine is memory. Samsung’s chips sit inside the servers that train and run AI models, and demand for them has outrun supply. The Next Web reports that demand for AI infrastructure has grown faster than the supply of memory chips, and that the gap is expected to last into 2027. High-bandwidth memory, or HBM, is the star product. These stacked chips sit beside the graphics processors that do the number-crunching and feed them data fast enough to keep expensive hardware busy. Douglas Research Advisory analyst Douglas Kim estimated that Samsung’s HBM shipments rose about 50 percent from the previous quarter, according to a report on the guidance. A preview from Crypto Briefing said revenue from HBM4, the newest generation, was expected to triple sequentially, though that was an analyst expectation made before the guidance, not a company figure.
Market share has moved too. Whalesbook reported that Samsung’s share in the specialized AI memory segment reached 33 percent in the second quarter of 2026, up from 21 percent earlier in the year. For years, Samsung trailed a rival in the race to supply HBM to Nvidia, and its recovery is a central part of the current story. Startup Fortune said Samsung’s latest HBM4 chips have cleared qualification with Nvidia and AMD, which would explain the improved shipments, though that is one outlet’s account and the company has not published qualification details in the material reviewed.
Prices help explain the profit. A shortage of memory pushes up what customers pay, and the rise extends beyond HBM to standard DRAM and NAND flash. SK Securities analyst Han Dong-hee estimated Samsung’s operating margin in its memory chip business at 76 percent for the quarter, unchanged from the previous one. That is an extraordinarily high figure for a manufacturing business, and it shows how much pricing power suppliers hold when demand exceeds supply. One report did raise a point worth watching, noting that memory prices may be starting to cool, which could shape how much of the demand reaches the bottom line over time.
How long the shortage lasts is disputed. Startup Fortune said Samsung expects the memory shortage to continue until 2028, while The Next Web reported a gap lasting into 2027, and a related headline in the same coverage said Micron sees the shortage tightening into 2028 after its own record quarter. Those timelines differ, and only the companies know what they are seeing in orders. Analysts quoted by TelecomLead expect fourth-quarter operating profit to grow about 8.2 percent from the third quarter, compared with roughly 20 percent growth in the third, which implies the pace of increase is slowing even if the level remains high.
The strength in chips masks weakness elsewhere. SammyGuru noted that Samsung did not disclose divisional figures, but that one division may post an operating loss of around 1.8 trillion won, and other coverage has contrasted the chip boom with losses in the foundry and mobile businesses. Samsung also makes smartphones and televisions, and TelecomLead pointed out that rising memory costs hurt phone makers, including Samsung itself, which buys components. So the group is both the biggest beneficiary of the shortage and a customer paying more for the same chips in its devices.
The wider implications are significant. Samsung’s results are a window into AI spending by the companies buying chips. When a technology group earns $80 billion in a quarter, it is because its customers, cloud providers, AI labs and chip designers, are spending enormous amounts on hardware, much of it financed by debt. Earlier this week, reports said SpaceX was in talks to borrow $40 billion to buy Nvidia chips, which illustrates the scale of capital flowing into the sector. Skeptics argue that this spending depends on future AI revenue that has yet to appear, while supporters say shortages show demand is real. The guidance supports the second view for now, because chips are being sold at high prices today, but it does not settle whether the buying will last.
For consumers, the picture is less cheerful. Higher memory prices tend to raise the cost of phones, laptops and other devices, and TelecomLead advised telecom operators planning cloud or edge projects to examine memory costs closely in equipment quotes. For Korea, a company contributing this much profit makes the economy more dependent on a single sector, a concern economists often raise about export-heavy markets. For investors, the key variables are how long the shortage lasts, how quickly rivals add capacity, and whether AI customers keep spending at the current pace.
Readers who follow global technology and business developments can find more coverage at BusinessTech Nigeria. The company publishes its announcements on its Samsung website, and The Next Web has a clear summary of the guidance in its report on the record profit. The full earnings release later this month will show the divisional breakdown, and with it a clearer view of how much of the $80 billion comes from memory alone.
