AI chip cycle to last through decade
The AI infrastructure build-out is in its early stages and is projected to run through 2030, supported by enterprise adoption and sovereign AI deployments.
The argument
The guest argued that despite potential periods of digestion, the supply-demand imbalance will persist because adding manufacturing capacity takes years and billions of dollars. Furthermore, key chipmakers currently trade at attractive valuations below market multiples.
The thesis, stress-tested
✓ What validates it
- ✓Sustained double-digit revenue growth for major hyperscalers
- ✓Enterprise AI software monetization metrics accelerating
▸ Risks discussed
- ▸Periods of inventory digestion or absorption
- ▸High capital requirements for manufacturing fabs
Hear it yourself
"There's a lot of deployment that, different, countries and states and regions, need to do to be, self sufficient. So, no, we are extremely, bullish, and I think that means very good things, across the semiconductor supply chain, whether one looks at, computing companies, whether one looks at networking, optics, memory. And, you know, these kind of deployments are never a straight line. There are are there are gonna be periods of, digestion in, between. But I think that, despite that, volatility, I do think the cycle can last till the end of the decade."
AFFILIATE LINK · ZORTIX MAY EARN A COMMISSION · NEVER A RECOMMENDATION TO TRADE