For the past several years, conversations about AI infrastructure have focused on acceleration of more GPUs, more compute, and more cloud. But now in 2026, the conversation is shifting in a meaningful way. The next phase of AI-driven growth is not about adding endlessly to what organizations have.

It’s about rethinking how everything works together.

AI will continue its rapid rise, and demand for compute will not slow. But the real inflection point ahead is not raw performance. It is governance, ROI, capacity and facility infrastructure optimization. Organizations are being pushed to extract more value from existing infrastructure while ensuring environments are ready for both existing operations and emerging AI demands.

This is where the idea of “hybrid” takes on new meaning.

Hybrid data centers are no longer just about blending on-prem and cloud. They are about environments where AI and conventional environments coexist efficiently, where power, cooling, and physical infrastructure are as strategic as software stacks, and where modernization does not always mean replacement.

Our latest Direction of Technology report shows that AI continues to be a top investment priority for technology leaders, while infrastructure and security readiness remain critical enablers of success. As AI adoption deepens, these pressures are forcing organizations to rethink how existing environments support new workloads.

Rather than expanding without constraint, many organizations are focusing on execution, operational discipline, and resilience. Many are prioritizing optimization over unchecked expansion.

Infrastructure Is Becoming a Team Sport

One of the most underappreciated shifts is who now plays a role in technology decisions.

As AI workloads strain power, cooling, and capacity, technology architects can no longer work in isolation. Collaboration with trades, builders, and facility experts is becoming essential as mechanical, electrical, and plumbing systems are now central to performance, sustainability, and scalability.

This convergence is creating a new ecosystem where IT architecture and physical design move together. From refreshing a single server to retrofitting entire facilities, organizations are asking deeper questions about how infrastructure supports long-term outcomes, not just immediate demand.

And this evolution is not limited to hyperscalers.

Private and smaller data centers are seeing renewed attention as organizations recognize the value of distributed, hybrid approaches. The goal is not to centralize everything, but to place workloads where they perform most efficiently within real-world infrastructure constraints.

New Markets, New Conversations

As infrastructure modernizes, the boundaries of where technology lives are expanding.

AI-driven workloads are moving into industries and environments that historically sat outside advanced IT conversations. Technology is being embedded into places we never imagined. Construction projects, logistics operations, manufacturing sites, and even recreational developments are becoming technology-enabled spaces. A project that once required little more than physical materials may now demand cameras, gateways, and edge servers to deliver real-time analytics and insight.

This evolution is changing who the buyers are, how solutions are sold, and what success looks like. The targets have changed, and relationship-driven sales are evolving as technology becomes part of everyday operations across industries previously untouched by complex IT.

Balance, Not Excess

The defining characteristic of 2026 will be balance.

Organizations that thrive will not be those that chase every new technology, but those that align innovation with governance, performance with sustainability, and ambition with operational reality. Hybrid data center strategies allow organizations to make what’s old new again, extend hardware lifecycles, reduce waste through circular practices, and modernize in ways that are financially and environmentally responsible.

What Comes Next

As organizations accelerate AI adoption, one constraint is becoming increasingly clear: power. But power is only part of the equation because as AI racks push density higher, organizations are also rethinking how they manage heat, exploring liquid cooling, enhanced HVAC strategies, and other alternative thermal methods designed to support next-generation workloads.

Industry conversations are beginning to acknowledge what many are calling the “AI Energy Wall”. The question is no longer only how many GPUs an organization can deploy. It is whether the infrastructure behind those systems can sustain the energy demands required to run them effectively.

Power availability, electricity costs, and infrastructure readiness are quickly becoming central considerations in modern IT strategy.

Through initiatives like the Hybrid IT Energy Initiative teams across the ecosystem are beginning to address this challenge directly. Teams across the ecosystem are focused on helping partners and customers evaluate whether their environments are ready and where modernization is needed in ways that balance performance, energy consumption, and long-term operational sustainability.

For many organizations, the first step is understanding where potential gaps exist today.

Recent industry research from HPE’s ProLiant Gen11 User Evidence Survey has found that nearly 50 percent of enterprise servers in operation today are Gen9 or earlier, infrastructure that predates many modern AI workloads. These legacy environments often require significantly more power and cooling than newer generations, highlighting both the scale of the modernization opportunity and the operational risks organizations face as AI demands continue to grow.

For partners and customers looking to take action, there are several ways to begin:

• Audit your environments to identify customers who have recently deployed GPUs or who are operating on aging infrastructure that may struggle to support modern AI workloads.
• Launch an energy or infrastructure readiness assessment to better understand potential power, cooling, and performance gaps before they become operational constraints.
• Engage with the Hybrid IT Energy team to explore modernization strategies and cross-supplier collaboration opportunities.

If you are interested in learning more or contributing ideas to this initiative, reach out to the Hybrid IT Energy team at hybriditenergy@tdsynnex.com to continue the conversation.

The future of AI will not only be defined by how powerful our systems become, but by how efficiently and sustainably we can power them.

Citation:

Source: HPE ProLiant Gen11 User Evidence Survey, June 2024 (Hewlett Packard Enterprise).