AI trade shifts to supply chain bottlenecks
The AI investment thesis is expanding beyond GPUs to critical physical bottlenecks along the data center supply chain, including cooling, electrification, power, and memory.
The argument
Sebastian Page of T. Rowe Price and Heath Terry of Citi Research argue that investors should focus on where the infrastructure money is being spent rather than the hyperscalers spending it. They highlight that while hyperscalers face margin pressure, bottleneck providers like memory and power companies possess immense pricing power due to irresistible demand meeting limited supply.
The thesis, stress-tested
✓ What validates it
- ✓Continued acceleration in Azure and cloud infrastructure revenues
- ✓Sustained high gross margins for memory manufacturers in upcoming quarterly reports
▸ Risks discussed
- ▸Hyperscaler free cash flows are declining rapidly
- ▸High valuations in software segments of tech giants
- ▸Potential cyclical normalization of memory margins
Hear it yourself
"From our global headquarters in New York City, we are live on Bloomberg Television weekday mornings from six to 9AM eastern. Subscribe to the podcast on Apple, Spotify, or anywhere else you listen. And as always, on the Bloomberg terminal and the Bloomberg business app. We begin this hour with stocks pushing higher following a boost from Micron's blowout earnings report. Sebastian Page of T. Rowe Price writing AI is an expanding theme from GPUs to other bottlenecks along the data center supply chain as Buzz Lightyear said to infinity and beyond. Sebastian joins us now for more."
AFFILIATE LINK · ZORTIX MAY EARN A COMMISSION · NEVER A RECOMMENDATION TO TRADE