Samsung’s profits jump 19x in a year and you don’t need AI to figure out why
Share price down sharply, apparently amid fears the bubble is getting bigger
systems Samsung has delivered guidance for its second quarter results and forecast profits 19 times higher than it managed in the same period of 2025.
In Q2 last year, the Korean giant’s sales totalled ₩75.7 trillion ($49.7 billion) and operating profit landed at ₩4.68 trillion ($3.1 billion).
Fast forward a year, and Samsung told investors that once its accountants finish their work, they’ll report sales of ₩170 trillion to ₩172 trillion ($111.9 billion) and operating profit between ₩89.3 trillion and ₩89.5 trillion ($58.4 billion).
Samsung’s guidance announcements are very brief, and don’t offer any explanation for its performance, but in its Q1 results announcement the company said its memory business set all time revenue and profit records and attributed that result to “technological leadership in the memory market, as well as higher average selling price.”
Neither of those factors have changed since, and the AI boom continues. Samsung has clearly made bank and has the chance to do so for another few years given memory prices show little sign of falling. Indeed, Samsung’s memory-making rival Micron recently found plenty of buyers willing to pay a premium for years to come.
Strangely, Samsung’s mountain of profit didn’t enthuse investors as its share price fell almost ten percent. Korea’s other big memory-maker, SK Hynix, also saw its scrip sink.
- Samsung floats 2028 launch for seaborne datacenter
- South Korea’s hot new sensation is 3S+1F – a quadrillion-Won AI plan, not a band
- SK Hynix to boost memory production 3x ... you can wait another 8 years, right?
- Expect more of those DRAM price hikes as memory shortage continues to bite
Analysts reportedly worry that the huge efforts to build new chipmaking plants could mean the world eventually has an oversupply of memory – especially if AI companies post results that signal a slowdown their infrastructure rollout plans. Such results would likely follow difficulties monetizing existing investment in AI infrastructure, perhaps because user experiences with the tech are often not great: The Register has reported attempts to use AI for customer service being wound back, complaints about the tech’s negative impact on CI/CD pipelines, and businesses plowing ahead with AI spend despite elusive ROI. ®