“The data center industry continues to rapidly evolve how it designs, builds, operates, and maintains data centers in response to the density and speed-of-deployment demands of AI factories,” said Scott Armul, chief product and technology officer at Vertiv. “We’re seeing forces across technologies, including extreme densification, driving transformative trends such as higher-voltage DC power architectures and advanced liquid cooling, which are critical to achieving the gigawatt-level scalability that is essential for AI innovation. On-site power generation and digital twin technology are also expected to contribute to advancing the scale and speed of AI adoption.”
The Vertiv Frontiers report builds upon Vertiv’s previous annual predictions of data center trends. The report identifies the macro forces driving data center innovation: extreme densification—accelerated by AI and HPC workloads; Scaling to gigawatt levels at breakneck speed—data centers are now being deployed rapidly and on an unprecedented scale; the data center as a computing unit—the AI era demands that facilities be built and operated as a single system; and silicon diversification—data center infrastructure must adapt to an ever-increasing range of chips and computing capabilities.
The report details how these macro forces have, in turn, shaped five key trends impacting specific areas of the data center landscape.
Power Boost for AI:
Most data centers today still rely on hybrid AC/DC power distribution from the grid to IT racks, which involves three to four conversion stages and some inefficiencies. This existing approach is under pressure as power densities increase, driven largely by AI workloads. Moving to higher-voltage DC architectures allows for significant reductions in current, conductor size, and the number of conversion stages, while centralizing power conversion at the room level. Hybrid AC/DC systems are widespread, but as full DC standards and equipment mature, higher-voltage DC is likely to become more prevalent as rack densities increase. On-site generation and microgrids will also drive the adoption of higher-voltage DC.
2. Distributed AI:
The billions of dollars invested to date in AI data centers to support large language models (LLMs) have been aimed at supporting the widespread adoption of AI tools by consumers and businesses. Vertiv believes that AI is becoming increasingly important for businesses, but how and from where these inference services are delivered will depend on each organization's specific requirements and conditions. While this will affect businesses of all types, highly regulated sectors, such as finance, defense, and healthcare, may need to maintain private or hybrid AI environments using on-premises data centers due to data residency, security, or latency requirements. Flexible and scalable high-density power and liquid-cooled systems could enable capacity through new construction or the retrofitting of existing facilities.
3. Energy Autonomy Accelerates:
On-site short-term power generation capabilities have been essential for decades for most standalone data centers to support resilience. However, widespread energy availability challenges are creating conditions for adopting expanded energy autonomy, especially in AI data centers. Investment in on-site power generation, using natural gas turbines and other technologies, offers several intrinsic benefits but is primarily driven by energy availability challenges. Technology strategies such as Bring Your Own Power (and Cooling) will likely be part of ongoing energy autonomy plans.
4. Design and Operations Driven by Digital Twins
With increasingly dense AI workloads and more powerful GPUs, there is also a need to deploy these complex AI factories rapidly. Using AI-based tools, data centers can be virtually mapped and specified through digital twins, and critical IT and digital infrastructure can be integrated, often as prefabricated modular designs, and deployed as compute units, reducing time-to-token by up to 50%. This approach will be key to efficiently achieving the gigawatt-scale deployments needed for future AI advancements.
5. Adaptive and Resilient Liquid Cooling
AI workloads and infrastructure have accelerated the adoption of liquid cooling. But conversely, AI can also be used to further refine and optimize liquid cooling solutions. Liquid cooling has become mission-critical for a growing number of operators, and AI could offer ways to further enhance its capabilities. AI, in combination with additional monitoring and control systems, has the potential to make liquid cooling systems smarter and even more robust by predicting potential failures and effectively managing fluids and components. This trend should translate into greater reliability and uptime for high-value hardware and the associated data and workloads.
Vertiv operates in more than 130 countries, providing critical digital infrastructure solutions to data centers, communications networks, and commercial and industrial facilities worldwide. The company's broad portfolio encompasses energy management, thermal management, and IT infrastructure solutions and services, from the cloud to the network edge. This integrated approach enables continuous operations, optimal performance, and scalable growth for customers facing an increasingly complex digital landscape.
