Understand any stock in simple English
Enter a ticker to learn what the company does and how it makes money.
SK hynix makes memory chips—the fast working memory in computers and phones and the long-term storage chips in modern devices—and sells them to the companies that build electronics.
SK hynix is one of the world's biggest makers of memory chips. Its two main products are DRAM, the fast short-term memory a computer uses while it's running programs, and NAND flash, the storage that keeps your photos and files safe when the power is off. It sells these chips to device makers and to the big cloud companies that stuff thousands of them into data centers. You almost never buy anything directly from SK hynix—it's the supplier behind the brands you know.
The business runs on a brutal cycle. Memory chips are a commodity: one company's DRAM works much like another's, so buyers mostly compete on price, and the market swings between severe shortages and gluts. When demand is hot, prices and profits soar; when everyone builds too much factory capacity at once, prices crash. SK hynix is one of only a handful of companies with the scale and technology to survive those swings—alongside Samsung and Micron—and that small club gives it real power. The most important recent twist is high-bandwidth memory, or HBM. AI systems need to feed data to their processors incredibly fast, and HBM is a stack of memory chips wired together to do exactly that; it's a major reason AI chips are so expensive. SK hynix became the leading supplier of HBM to Nvidia, the dominant AI chip designer, and that position turned a cyclical commodity maker into a key player in the AI boom. The risk is real, though: rivals are racing to catch up, and memory prices can still collapse. SK hynix spends heavily on new factories and chip development—memory is a capital-hungry business where falling behind on technology is fatal. It also pays dividends and buys back shares. The company is headquartered in Icheon, South Korea, close to Seoul, and its fate is tightly tied to the chip cycle and, increasingly, to how much the world keeps spending on AI.
Saved answer · last researched 21 Oct 2026
see a list of the most recent 300 searches below: