Avatar photo
Written by

Enrique Ortegon

28 Jan, 2026 4 minutes

Reflections after PTC26 on power, planning cycles, and what AI is changing for digital infrastructure.

PTC has long been the most important industry event to start the year. It’s where new projects are announced, strategies are tested, and early signals become clearer. PTC26 was no exception.

One idea came through consistently in public discussions, panels, and coverage:

The old infrastructure playbook is obsolete, not because it was wrong, but because AI has changed the physics of demand.

Current AI workloads are forcing changes in how data centers, networks, and energy systems are planned and delivered.

Power now sets the starting point

For years, power has been a growing concern in data center planning. What has changed is its position in the decision process. Power availability is no longer something to validate after site selection; it increasingly defines where infrastructure can be built in the first place.

Public analysis from the International Energy Agency and the World Economic Forum shows electricity demand from data centers accelerating sharply, driven largely by AI workloads. At PTC26, this shift was reinforced by a recurring theme across industry coverage: while capital and compute can scale quickly, power systems, permitting, and grid expansion move at a very different pace.

For data centers and network planning, this shifts the sequence of decisions. Infrastructure increasingly forms around locations where power can be delivered at scale, and networks are then designed to support growth around those anchors. In this context, timelines matter less than reliability. Longer delivery cycles can be acceptable; missed commitments are not.

AI is compressing planning cycles

The second shift is about timing.

Traditional infrastructure planning relied on relatively smooth demand curves. Capacity could be phased in based on forecasts that were directionally correct. AI workloads are challenging that model.

Public reporting around PTC26 highlighted how AI deployment is outpacing the physical systems that support it. Compute scales fast. Power grids, permitting, and construction do not. The result is not instability, but compression. Decisions happen earlier, requirements evolve faster, and precision is harder to achieve upfront.

For interconnection and network planning, this means systems must be designed for change:

  • Network infrastructure needs clear expansion paths.
  • Interconnection hubs become critical points for diversity and resilience in data transit.
  • Infrastructure must respond to bursty, fast-changing demand, not assume predictable growth.

In this environment, the value of an interconnection data center lies first in the density of its ecosystem: the concentration of digital network operators and enterprises already exchanging traffic. Capacity matters because it allows that ecosystem to grow as needs evolve, without fragmenting traffic or performance.

Locality and control are now design inputs

Data sovereignty increasingly informs where data centers are built, how networks are deployed, and how traffic and interconnection are engineered.

The World Economic Forum has written about AI’s distributed future, where workloads balance centralized training with localized processing. Economist Impact has highlighted how sovereignty concerns affect trust, risk management, and operating models.

For infrastructure decisions, the takeaway is practical:

  • Where data is processed matters.
  • How traffic is routed matters.
  • Locality and control influence architecture choices, not just compliance.

AI amplifies this dynamic. As workloads become more distributed, regional hubs and dense interconnection points become more important. Network design has to support locality without sacrificing performance or flexibility.

What PTC26 made clear

PTC26 didn’t just surface these trends. It reinforced how tightly they are converging.

Power availability, compressed planning cycles, and locality are no longer separate conversations. Together, they are redefining how digital infrastructure is planned and built.

The old playbook assumed predictability. The next one assumes adaptability.

Additional Resources

The following articles and reports informed the perspectives shared in this post and provide further context on how AI is reshaping digital infrastructure, power planning, and network architecture:

Avatar photo
Enrique Ortegon

Popular tags