Year-end reflections: scale, speed, and new technical expectations
A year defined by surprising scale and fast shifts
At the start of 2025, many of us expected a normalization year for the energy sector. Instead, within weeks, the market changed direction. Demand patterns shifted rapidly, driven primarily by data centers, and the implications reached far beyond order intake.
This year forced a reassessment of how turbomachinery is produced, supported, and digitally enabled. The scale and speed of change exposed gaps in traditional service models and highlighted where new technical approaches are required.
Data centers: the most significant market shift of 2025
The most notable change in 2025 was the emergence of data centers as a major turbomachinery customer segment. Orders for the NovaLT™16 reached approximately 500 units in a single year. For comparison, fewer than 50 units of this model were ordered across the previous decade.
This shift has immediate consequences. Manufacturing facilities designed to produce roughly 20 units per year must now scale output by a factor of three to four and, in some cases, more. That challenge is not limited to production capacity. It affects testing, commissioning, spare parts planning, and long-term fleet support.
Data center developers are not traditional energy operators. They are digital-native organizations with expectations shaped by automation, real-time monitoring, and high service responsiveness. For this fleet, legacy service approaches are not sufficient. The service model must be designed digital-first, starting at installation and continuing throughout the lifecycle.
Aha moment #1: digital enablement must be designed in
Data center operators run AI models as their core business. Power generation is a means to enable compute, not an isolated utility function. It is inconsistent for a technology provider to supply power for AI workloads without providing a digitally integrated product and service model.
For this segment, digital capabilities cannot be added later. Remote monitoring, unified data pipelines, and AI-ready analytics must be foundational elements of the turbine architecture. Commissioning, performance validation, and predictive maintenance all need to be designed around continuous data availability and integration with customer systems.
This represents a shift toward what is effectively a digital-native turbine fleet, deployed at scale for the first time.
Aha moment #2: customers are opening their digital ecosystems
Another significant development in 2025 was the increased openness of customers regarding their digital platforms and strategies. Historically, many operators invested heavily in digital programs but kept them siloed, with limited external integration.
That dynamic is changing. Customers are now actively requesting collaboration around interoperability between their platforms and iCenter, joint solution architecture, and co-development of workflows for automation and optimization.
This shift requires a different engagement model. Technical discussions now extend beyond equipment performance to include APIs, data ownership, cybersecurity boundaries, and system interoperability. For Baker Hughes, this places greater emphasis on integration readiness and a more consulting-oriented approach, particularly for solution architects.
AI and energy: a two-way street
Artificial intelligence sits at the center of this transformation. AI workloads require large, reliable, and efficient power supplies, which is driving demand for high-efficiency gas turbines in data-center applications.
At the same time, AI is increasingly used to optimize energy systems themselves. Advanced analytics can improve fuel efficiency, manage thermal behavior, and enhance overall operational performance.
This creates a two-way relationship:
- Power for AI, where turbines such as the NovaLT16 support large-scale compute.
- AI for power, where analytics and machine learning optimize turbine and plant performance.
The result is a tighter coupling between energy technology providers and digital operators, both technically and operationally.
Implications for Baker Hughes and solution architects
The implications are clear. Technical readiness and integration capability are becoming central differentiators. Products must be ready for rapid digital integration with customer platforms. Interoperability between iCenter and external systems must continue to advance.
Solution architects sit at the intersection of hardware, software, and customer workflows. Their role is increasingly strategic, translating operational requirements into integrated digital architectures. Collaborative models are no longer optional. They are essential to accessing opportunities that were previously constrained by closed systems.
Looking ahead: from analysis to action at the Annual Meeting
Our Annual Meeting in Florence will be an opportunity to move from trend analysis to practical solution-building. Discussions will focus on how to support the lifecycle of a new turbine fleet profile, how to integrate digital architectures between Baker Hughes and digital-native customers, and how to address energy efficiency challenges in AI-intensive environments.
The objective is alignment: shared priorities, clear development paths, and concrete next steps.
A year that reset expectations
2025 reset expectations for energy technology providers. Digital-native customers, AI-driven workloads, and expanding collaboration models are now part of the operating environment.