Andy Jassy disclosed in his annual shareholder letter that Amazon’s AI services at AWS are running at a $15 billion annual revenue rate. He also revealed that the company’s custom chip business — Graviton and Trainium — has crossed a $20 billion annualized run rate, roughly double what the company cited earlier this year.

Amazon stock jumped more than 5% on the news. The S&P closed at 6,824 and the Nasdaq at 22,822 on Thursday as the letter gave investors a concrete number to attach to years of capital expenditure narrative.

Samsung reported that first-quarter profit likely surged more than 700% year-over-year, driven by explosive demand for high-bandwidth memory chips used in AI training and inference systems. The number is staggering but follows a period when Samsung’s memory business was being undercut by inventory gluts and pricing pressure.

The reversal is a clean signal: the AI buildout has moved deep enough into the stack that it is now pulling on memory, not just logic chips. Nvidia gets most of the narrative oxygen when people talk about AI hardware, but the demand chain runs straight through DRAM and HBM suppliers. Samsung, SK Hynix, and Micron are all beneficiaries of the same underlying dynamic.