Key insights
- The author discusses the rise of Micron ($MU) as a key AI beneficiary due to HBM demand and pricing power. Having sold $MU too early, they are now exploring other potential "obvious in hindsight" AI-related investments, including storage ($WDC, $STX), networking ($AVGO, $MRVL), power, cooling, and data center infrastructure. This suggests continued investor interest and potential rotation into other chip-adjacent sectors.
Micron turning into a trillion-dollar AI winner still feels crazy.
A year ago most people were focused on GPUs, hyperscalers and software. Now memory is suddenly the bottleneck everyone cares about. HBM demand is tight, pricing power is back, and the whole thesis looks stronger than it did when the stock was much cheaper.
I was in $MU earlier, took profits, and completely underestimated how big the AI memory trade could get. Classic mistake: I saw the thesis, but not the full rerate.
Now I’m trying to figure out where the next “obvious in hindsight” move might be.
Is it still memory? Is it storage? Is it networking? Is it power? Is it cooling? Is it data center infrastructure?
Names I’m looking at: $WDC, $STX, $AVGO, $MRVL, $ALAB, $CRWV, $NBIS.
Anyone else sell $MU too early?
And what do you think is the next AI bottleneck trade?