The memory shortage may be eating into IT buyers' budgets and forcing reworks of infrastructure buildout plans, but the component makers look to be riding the crisis all the way to the bank.

Both Micron and SK Hynix – two of the so-called "big three" memory makers – saw their market caps pass the $1 trillion mark this week as their stock prices surged amid unprecedented demand.

Micron topped $1 trillion for the first time earlier this week after UBS tripled its price target on the firm’s stock from $535 to $1,625. The investment giant cited long-term agreement opportunities, with AI chip firms lining up to snap up its high-bandwidth memory (HBM) lines.

The Boise, Idaho-based firm had been a volatile stock for some time, though the AI boom has eased some of its historical swings. It did see a significant dip in late March, coinciding with a Google compression algorithm dubbed TurboQuant going viral, with online users speculating on the tool’s ability to shrink weights for use in long-context AI inference.

Micron even packed up its bags in the consumer space to follow the money and double down on AI data center customers, a move that has already paid dividends, with revenue for the firm’s second quarter (Q2) of fiscal 2026 coming in at $23.86 billion, almost triple its $8.05 billion from a year prior.

SK Hynix, meanwhile, joined the $1 trillion market cap club after a similar stock surge. The company’s stock price is up 231% for the year so far, with the fellow big three member posting record profits.

The demand for memory hardware has seen the South Korean firm admit lead times are being pushed out as far as 2028, with HBM and small outline compression attached memory module (SOCAMM) lines largely affected.

SK Hynix has sought to consolidate its position by establishing low-power wafer factories to address supply chain issues while also spending $10 billion on an AI-focused carve-out with its U.S.-based Solidigm arm. It’s also significantly raised prices amid soaring demand, while also making “strategic investments” in Spanish RISC-V chip company Semidynamics to advance “memory-centric AI inference architecture.”