South Korea's two memory-chip champions are placing an enormous bet on the artificial-intelligence boom, pledging more than $550 billion to expand production in a drive coordinated with the government, TechCrunch reported.
The commitment
Samsung Electronics and SK Hynix said the money would fund new memory-chip fabrication plants and additional capacity for high-bandwidth memory (HBM), the specialized chips that sit alongside the processors that train and run AI systems. The plan was unveiled with the backing of the South Korean government, which has framed semiconductors and AI as central to the country's economic future. It forms part of a far larger, longer-term investment push by the two firms — newsparlor reported earlier this week on a roughly $1.3 trillion decade-long capital plan that had unsettled some investors with its sheer scale.
What 'RAMageddon' means
The spending is a reaction to a genuine crunch. "RAMageddon" is the industry's tongue-in-cheek label for a worldwide shortage of memory chips — both the DRAM found in everyday computers and phones and, more acutely, the premium high-bandwidth memory used in AI accelerators such as Nvidia's. As technology companies race to build out AI infrastructure, demand for memory has outstripped supply, sending prices sharply higher and rippling through the broader electronics market, raising costs and, in some cases, delaying products.
Why it matters
Few companies are as central to this story as these two. Samsung and SK Hynix together account for roughly 70% of the world's DRAM, and SK Hynix in particular has become the leading supplier of HBM for cutting-edge AI chips. When their output is constrained, the effects are felt everywhere from laptops to the giant data centers underpinning the AI build-out. For South Korea, memory chips are among the most valuable exports, making the crunch — and the response — a national economic priority as well as a corporate one.
A familiar gamble
For all the confidence on display, the memory business is famous for its booms and busts. Time and again, shortages have driven prices up and lured manufacturers into pouring money into new plants, only for the resulting glut to send prices tumbling and profits with them. The latest mega-investment rests on a conviction that this cycle will be different — that AI demand is large and durable enough to absorb the new capacity. If that bet is right, the build-out will help relieve a bottleneck choking the AI era. If demand cools, the industry could find itself, once again, with far more memory than the world wants to buy. None of this is investment advice.



