
Samsung has posted the kind of quarter that shows just how much the artificial intelligence boom is reshaping the old hierarchy of the semiconductor market. The company is still a consumer electronics giant, but right now its biggest story is memory, servers and the shortage of advanced chips needed to keep AI data centres growing.
In its second-quarter 2026 results, Samsung said consolidated revenue reached KRW 171.5 trillion, another all-time quarterly high and up 28 percent from the previous quarter. Operating profit rose to KRW 89.5 trillion, also a record, while earnings per share for common and preferred shares increased 52 percent to KRW 10,849.
The engine behind that performance was the Device Solutions division, which includes Samsung’s semiconductor operations. The division posted KRW 127.5 trillion in consolidated revenue and KRW 89.2 trillion in operating profit for the quarter. Samsung said its Memory Business achieved another record quarter by prioritising AI server demand, especially high-value products that are in short supply across the industry.
That is the clearest signal from the results. Samsung is benefiting from the same AI infrastructure cycle that has lifted memory rivals, cloud suppliers and storage companies. Server DRAM, enterprise SSDs and high-bandwidth memory are becoming strategic products because AI models need huge amounts of fast memory close to compute. Samsung said it has scaled HBM4 sales and shipped HBM4E samples to major customers, which puts it deeper into the next round of AI hardware competition.
The company also expects the second half of 2026 to remain strong, driven by server demand, continued AI infrastructure capital expenditure and wider adoption of agentic AI. It said supply constraints are likely to continue despite efforts to increase production. That point matters because an undersupplied market gives memory makers pricing power, but it also raises costs for companies building AI servers, smartphones, PCs and enterprise hardware.
This is why the memory boom has become one of the most important parts of the AI economy. Earlier this week, SK Hynix posted record AI memory profit, but investors still worried about expectations and future spending. Samsung’s numbers now show that the opportunity is broad, but so is the pressure. Everyone wants more AI memory, and everyone is trying to expand capacity without overspending into the next cycle.
Samsung’s foundry business also improved, helped by HBM base-die demand and stronger orders from U.S. customers. The company pointed to more 2nm high-performance computing engagements and plans to ramp production of new mobile products based on its second-generation 2nm process in the second half. That is important because Samsung is still trying to strengthen its foundry position against TSMC, whose own record quarter was also powered by AI chip demand.
The consumer side of Samsung looked less clean. The MX and Networks businesses posted KRW 33.2 trillion in revenue but recorded an operating loss of KRW 0.7 trillion. Samsung said sales of the Galaxy S26 series and Galaxy A series were solid, but component costs weighed on earnings. The company is now looking to the Galaxy Z8 series, AI features and new form factors such as intelligent eyewear to improve its premium mix, following the company’s recent Galaxy Z Fold8 and Flip8 launch.
That contrast is useful. Samsung is making record money from the infrastructure powering AI, but its own devices are also feeling the cost of the same supply chain. Memory prices help the semiconductor division, while higher component costs can hurt phones, displays and consumer electronics. Few companies sit on both sides of that equation as visibly as Samsung.
The quarter makes Samsung one of the clearest winners of the current AI hardware cycle. The risk is that memory and foundry markets can turn quickly if supply catches up too fast or if cloud companies slow spending. But for now, Samsung has what investors want to see from a major tech earnings report: record revenue, record operating profit, direct exposure to AI infrastructure and a stronger claim to the next generation of memory technology.







