AI is pushing the grid to its limits — and transformers are now the critical path for both data centers and renewable projects.

Today’s guest, Doug Banty, EVP at Forgent (and former CEO of MGM Transformers), sits at the intersection of these two worlds. Half of his week at RE+ was spent across the strip at the Yotta data center conference, where the only topic on everyone’s mind was power. In this episode, Doug explains how AI-driven rack density is rewriting the entire powertrain, why hyperscalers are treating speed like an existential priority, and what this means for developers trying to keep pace.

“Every Nvidia chip has higher and higher wattage at the rack. We're looking at a kilowatt at the rack very soon. So every time you do that, you have to effectively reinvent the entire powertrain.” – Doug Banty

He also breaks down the industrial logic behind Forgent — a merger of MGM, VanTran, States, and PowerCube — and how integrated solutions are becoming essential for execution in an era where timelines matter more than price.

Expect to learn:
⚡ Why data centers are moving at crisis-level urgency
⚡ The difference between power and distribution transformers
⚡ What MGM’s 500,000 sq ft Waco factory means for lead times
⚡ How Forgent reduces execution risk through integrated packages
⚡ The single most effective way developers can avoid delays

Hit play — and see what the future of the grid really looks like.

RESOURCES:

Connect with Doug Banty:

• LinkedIn

Check out MGM Transformers:

• Website | LinkedIn

Check out Forgent Power Solutions:

• ‍Website | LinkedIn

‍

🕒 TIMESTAMPS/Chapters

00:00 — Welcome + AI power demand framing

01:07 — MGM/VanTran history and capabilities

03:38 — Why Forgent was formed (4-company merger)

05:15 — Non-obvious advantages for customers (execution, risk reduction)

06:19 — Are transformers still the bottleneck? (nuanced answer)

07:58 — MGM’s 500,000 sq ft Waco facility ramping output

09:12 — Data center vs. renewables urgency

10:35 — Why AI demand is an existential challenge for hyperscalers

12:05 — How powertrain design is changing

14:34 — What developers need to know about speed and execution

17:24 — Rack density and rapid engineering cycles

18:52 — How developers can save weeks on their project timelines

20:29 — Why early partnership unlocks allocation

20:45 — Closing thoughts

‍