Samsung Electronics said on 7 July that second-quarter operating profit was set to jump 19-fold, helped by demand for memory used in artificial-intelligence systems. Reuters reported that investors nevertheless wiped more than $80 billion from the company’s market value as they questioned how long the AI bonanza could last. [1]
Strong demand, fragile expectations
The contradiction is instructive. A company can report spectacular growth while its share price falls because the market was expecting even more, or because investors fear that customers will slow capital spending. Memory chips sit at the centre of the AI infrastructure build-out, but they are also cyclical products whose prices and orders can turn quickly.
The Semiconductor Industry Association said global chip sales reached $120.6 billion in May 2026, up 9.2% from April and more than double May 2025. [2] That data supports a strong demand story. It does not settle the valuation question: sales can rise while margins, mix and future orders become harder to forecast.
The supply chain is broader than one winner
Tom’s Hardware, citing Samsung management, reported that the semiconductor division expected 2026 profits to exceed its entire historical total from roughly four decades of chip operations. [3] The detail illustrates how concentrated the current cycle is in high-bandwidth memory and data-centre demand. It also shows why foundry, logic and handset weakness can matter even when memory is tight.
What to watch
The next signals are customer capex, memory contract prices, inventories and the split between high-margin AI parts and slower legacy products. The lesson for readers is not that the AI boom is over. It is that a real technology build-out can still generate financial volatility when expectations outrun the information available about the next cycle.
The market lesson is to separate operating performance from expectations. Profit growth answers what the company earned; the share-price reaction reveals what investors had already priced in and what they fear about future orders. Memory demand, customer concentration, inventories and capital spending therefore matter alongside the headline multiple. [1][2][3]
