Episode summary
AI is changing the economics of enterprise infrastructure faster than many organizations expected. In this episode of Beyond the IT Headlines, Shawn Rosemarin talks with Bloomberg Intelligence analyst Woo Jin Ho about what is driving that shift, from hyperscaler demand and tighter storage capacity to rising power costs, cloud economics, and the growing appeal of consumption-based models. They explore why this cycle feels different, what enterprises may be underestimating about energy and infrastructure planning, and why more leaders may start thinking about data less like hardware and more…
Chapters
- 0:00 — The new economics of AI infrastructure
- 1:30 — Why this NAND cycle feels different
- 3:20 — Hyperscaler demand and five-year capacity commitments
- 4:45 — Will new supply in 2027 actually solve the problem?
- 5:50 — The edge, device refresh cycles, and delayed demand
- 7:00 — Why supply-chain resilience matters more than price
- 8:30 — The energy reckoning for enterprise IT
- 10:29 — Power access, colocation pressure, and rising cloud bills
- 12:15 — Tape, disk, and flash in a constrained market
- 14:55 — Why there’s no quick fix for the power bottleneck
- 17:20 — Data as a utility
- 19:43 — What true utility infrastructure should look like
- 22:00 — The wait-and-see camp vs. the innovator camp
- 25:05 — What Wall Street is signaling about infrastructure spend
- 27:44 — The biggest mistake customers may make in the next 18 months
Leave a Reply