Jun. 27 at 7:09 PM
Just saw this:
A new BofA report is floating the idea that close to
$1 TRILLION could go into memory by 2030.
That number is honestly hard to ignore.
To me it basically reinforces a few things:
Memory is becoming structural, not just cyclical
AI buildout is still early in terms of infrastructure
HBM/DRAM sit at the center of the stack no matter who wins
This doesn’t feel like a normal “upgrade cycle” anymore.
If even part of this plays out, names like
$MU aren’t trading a short-term cycle… they’re riding a multi-year demand wave.
Still watching this space closely. 👀
$DRAM