Analysts expect about 86 trillion won in operating profit for the second quarter, a third consecutive record, as DRAM and NAND rise. Samsung Electronics is on track for one of the strangest profit comparisons in its history. Analysts expect the company to report operating profit of about 86 trillion won for the second quarter, about $56 billion, when it releases preliminary guidance this week.
See also: iPhone Fold and MacBook Pro with touchscreen rumored again to have Samsung displays
That would be about 18 times the 4.6 trillion won it managed in the same quarter last year, when the memory business was still a burden rather than an engine.
The gap says less about a sudden turnaround and more about how far the AI memory cycle has come since Samsung passed $1 trillion in market value earlier this year. The number is an estimate, not a company number, coming from analysts polled ahead of the guidance, and the usual caveats apply. However, preliminary earnings in Korea rarely deviate much from reality, and the direction is not in doubt.
If achieved, it would be the third consecutive quarter of record operating profits for Samsung, each of them driven by the same thing: memory chips selling for much more than they did last year. Prices are the story. Citi Research put DRAM contract price growth quarter-over-quarter at about 44 percent, with NAND flash up about 53 percent, and HSBC reached similar conclusions. These are not the gradual moves the memory market is used to.
They reflect a shortage that has stopped looking cyclical and is starting to look structural, as demand for AI infrastructure continues to outpace what the big three memory makers can supply. Micron’s quadrupling revenue told the same story from the U.S. side of the industry. What has changed this cycle is scope. High-bandwidth memory, the stacked DRAM that sits next to Nvidia’s accelerators, remains the most profitable corner of the business, and Samsung has spent the last year trying to close the gap with SK Hynix there.
See also: Anthropic and Samsung discuss custom AI chip manufacturing
But the price hike now extends far beyond HBM. Standard DRAM and conventional NAND are also going up as AI workloads spill out of training clusters and into the usual server and storage architectures that support induction scale. This has implications far beyond the data center.
Memory makers have largely reallocated existing capacity to AI rather than adding new lines, so the same shortages that are boosting Samsung’s profits are also squeezing inventory for parts for phones and laptops. Apple felt it firsthand when it pulled the $599 Mac mini amid the DRAM crisis. Samsung is on both sides of this trade, selling off the scarce memory while its own device division pays more for it.
Samsung supplies memory to most of the companies building big AI systems, including Nvidia, Google, and Apple, which puts it near the center of the infrastructure chain even if it’s not the market leader. The competitive picture isn’t entirely comfortable, though. SK Hynix has moved faster on the latest HBM, locking in a multi-year supply deal with Nvidia, and the two Korean giants now trade the title of the country’s most valuable company with each other. There’s also the question of what such a large number of people are doing inside the company.
Samsung’s semiconductor division has generated the vast majority of the group’s profits, an imbalance that has already fueled a heated debate over how to share profits with employees. Record earnings tend to exacerbate those disputes rather than resolve them. The preliminary guidance will provide headline earnings and revenue figures but not segment-by-segment breakdowns, which come with full results later in the month.
See also: Samsung KNOX: Eight-year-old vulnerability put Galaxy devices at risk
Until then, the estimate stands as an indicator of how completely the memory business has been reshaped by a single demand cycle. Eighteen times last year's earnings is not a recovery. It's a
