SK hynix reported a 632% surge in third-quarter operating profit to $62.8 billion on explosive demand for AI-driven semiconductors, lifting Asian tech stocks and global investor confidence.
The South Korean chipmaker’s record earnings, its sixth straight quarter of revenue and profit growth, reignited optimism in the AI trade after a recent wobble over heavy capital spending and profitability concerns. Shares in SK hynix soared nearly 9% on Wednesday, while rival Samsung added over 8%, pushing Seoul’s Kospi index up more than 6%. Japan’s Nikkei 225 climbed over 1%, with chipmaker Kioxia surging more than 10%, Advantest up 3%, and Tokyo Electron 4%. Taipei also rose sharply on TSMC’s strong performance, with Hong Kong, Shanghai, Sydney, Singapore, and Wellington following suit.
The gains tracked advances in New York, where the Nasdaq added over 1% ahead of earnings reports from Microsoft and Alphabet later in the day. Analysts noted the results reinforced the AI supply chain’s long-term potential. “While Wall Street argues over whether the AI arms race can keep financing itself, SK hynix has just reminded the market that somebody on the other side of all that capex is receiving the cheque,” wrote Stephen Innes at SPI Asset Management.
The momentum followed Anthropic’s IPO filing on Monday, which revealed first-half revenue surged about 1,300%, and Broadcom’s announcement on Tuesday of a $63 billion financing structure with Citi, Blackstone, and Morgan Stanley to build new AI capacity. “The AI trade has undoubtedly received a healthy shot in the arm from Anthropic’s filing, with its incredible revenue growth and substantial commitments to future infrastructure investment,” said Pepperstone’s Chris Weston.
SK hynix, a leader in high-bandwidth memory chips for AI applications, said demand for its advanced semiconductors would outstrip supply until at least 2030. The company’s market capitalization has more than doubled since late 2025, reaching around $660 billion. Its key customers include Nvidia, which relies on its chips for AI infrastructure.
