Our own Emilio Vicens shared some thoughts with Pipeline Magazine about the AI energy roadmap and why a “gas-to-nuclear” bridge is the only realistic path to securing a zero-carbon future for the digital economy. Below is a brief excerpt from the article.

The AI revolution has collided with a stark physical reality: The U.S. power grid cannot support it. With data center power demand projected to nearly triple to over 130 gigawatts by 2030, the energy bottleneck has emerged as the primary threat to the AI roadmap. Tech leaders are now facing a new mandate: build, bring, or buy your own energy, or risk stalling your innovation entirely.

The Necessity of the “Bridge” for AI Data Centers 
The challenge is no longer just about capacity; it’s about timing. Grid interconnection timelines across the U.S. have stretched to three to seven years, far exceeding the 18 to 24 months it takes to build a modern data center. In some major hubs, like Columbus, Ohio, those timelines reach up to 84 months. This “time-to-power” gap is why 20% of planned data center projects are at risk of being canceled entirely.

Structuring the Transition from Gas to Nuclear Thermal 

While hyperscalers are making massive bets on Small Modular Reactors (SMRs), commercial scale is still a decade away. To bridge the gap, behind-the-meter gas generation offers a faster path to energization — often in as little as 16 to 30 months. 

Importantly, this is not a step away from sustainability. This flexible gas infrastructure creates the immediate capacity needed today while serving as a high-value resiliency layer once nuclear power comes online. By controlling their own generation, developers can bypass utility delays and secure their energy destiny.

We are excited to share our framework for navigating this transition and helping the industry move toward a more resilient, zero-carbon future. 

Read the full article here: Pipeline Magazine