Continental Breakfast with Exhibitors

DCF Editor in Chief Matt Vincent opens the Summit with a fast calibration of where the industry really is in 2026: from last year’s rapid expansion and “AI-ready” positioning to today’s hard reality of power, capital, labor, and community constraints converging at once. Rather than a traditional keynote, this session immediately turns to the room. Members of the Data Center Frontier Editorial Advisory Board will stand and deliver rapid-fire, two-to-three-minute perspectives from where they sit in the ecosystem—operators, developers, capital, grid, and community—to answer one central question: What has actually changed in the last year when it comes to building and delivering data center infrastructure, and where is execution breaking down or accelerating?

As AI workloads reshape infrastructure requirements, competitive advantage increasingly depends on access to power, entitled land, and the ability to execute at speed. The next wave of data center development is being driven not only by demand for compute, but by the strategic decisions that determine where AI capacity can be built and how quickly it can be delivered.

In this opening keynote fireside chat, EdgeCore Digital Infrastructure CEO Lee Kestler sits down with Data Center Frontier Founder and “Data Center Richness” Creator Rich Miller, in a conversation moderated by Data Center Frontier Editor in Chief Matt Vincent. Drawing on Kestler’s 25 years of experience serving hyperscale and enterprise customers—including leadership roles at EdgeCore, Vantage Data Centers, and DuPont Fabros Technology—the discussion will examine how leading developers are selecting markets, partnering with utilities, and planning for the next generation of AI growth.

Together, they will explore:

• How access to power and entitled land is reshaping data center site selection and long-term development strategies
• Why execution speed has become a defining competitive advantage in the AI era
• How developers are collaborating with utilities, governments, and customers to accelerate deployment
• The evolving dynamics driving hyperscale expansion and enterprise AI adoption
• What today’s infrastructure leaders see as the biggest opportunities and constraints for future AI growth

Providing an insider’s view into the forces redefining data center expansion, this keynote sets the tone for the Summit by examining how AI infrastructure is actually being planned, built, and scaled in an era where execution has become as important as technology itself.

As AI factories take shape, infrastructure design must evolve in lockstep with GPU innovation. This panel brings together industry leaders to explore how next-generation AI environments are co-created—from workload requirements to simulation, power architecture, liquid cooling, and modular deployment. Learn how integrated design, digital modeling, and ecosystem collaboration are enabling faster, scalable, and efficient buildout of high-density GPU data centers to meet accelerating AI demand.

Power has become the gating factor for AI infrastructure, and the industry is responding by fundamentally rewriting how data centers are planned, powered, and delivered.

Interconnection delays, grid congestion, and speculative queue dynamics are forcing a shift away from traditional utility-dependent models toward power-first development strategies. In response, developers and operators are assembling a new energy stack combining on-site generation, combustion-free baseload technologies, large-scale battery storage, and advanced cooling architectures, the better to accelerate timelines and ensure operational certainty. What’s emerging is not a single solution, but a coordinated system. Power, cooling, and compute are no longer designed independently; they’re being engineered together as an integrated architecture capable of supporting ultra-high-density AI workloads. At the leading edge, this means rethinking not just how power is delivered, but how it is consumed. New approaches to system design, thermal management, and infrastructure integration are reshaping what “power-first” actually looks like in practice.

This panel brings together leaders across on-site generation, fuel cells, energy storage, and next-generation data center architecture to examine how projects are being built in spite of grid constraints, and how power and infrastructure are being co-designed to deliver capacity at AI scale. The discussion will explore how these systems come together in practice, where they succeed, and what it takes to make them repeatable.

Networking Lunch Break with Exhibitors

AI infrastructure is no longer a single design problem. It’s a spectrum.At one end are hyperscale “AI factories”—massive, tightly integrated systems aligned with next-generation platforms like NVIDIA’s Vera Rubin roadmap, pushing toward extreme rack densities and tightly coupled power, cooling, and compute architectures. At the other are smaller-scale enterprise and edge AI deployments, where inference, latency, and cost constraints often favor lower-density configurations that can be deployed flexibly in existing environments. Between these poles, the industry is being pulled in multiple directions at once. The roadmap suggests 100 kW racks are arriving, 200–300 kW is being engineered, and some projections point to 1 MW+ per rack—yet sustained production environments above 50–60 kW remain limited, and each step upward multiplies challenges across:

• Power delivery and distribution (including emerging 800 VDC and DC architectures)

• Liquid cooling adoption, integration, and serviceability

• Structural design and white space planning

• Operations, safety, and maintainability at scale This panel examines where density is actually landing in production today, and how different classes of AI workloads—training vs. inference, centralized vs. distributed—are shaping real-world design decisions. Topics include:

• When high-density is essential, and when it’s an over-optimization

• Where retrofits still work—and where only purpose-built high-density sites make sense • How emerging electrical architectures and rack-level designs are influencing new builds

• Whether the market will converge on a dominant high-density model, or fragment into multiple architectures aligned to different AI workload classes

Details coming soon!

Networking Break with Exhibitors

AI infrastructure announcements are everywhere. Delivery is not.

As demand accelerates, a growing gap is emerging between what gets announced and what actually gets built. Power constraints, interconnection delays, permitting friction, and shifting commercial realities are forcing developers to navigate a far more complex path from concept to capacity.

The challenge is no longer just securing land or raising capital. It is aligning execution across power availability, customer commitments, infrastructure readiness, and network architecture and interconnection strategy—while managing risk in an environment where timelines are uncertain and failure is costly.

This panel brings together leaders across development, commercial strategy, interconnection, and infrastructure risk to examine where projects are moving forward, where they are stalling, and what ultimately determines whether an AI data center makes it from announcement to energized capacity.

Networking Reception with Exhibitors

Continental Breakfast with Exhibitors

AI data centers are no longer assembled; they’re designed as integrated systems. Using simulation, digital twins, and GPU-accelerated modeling, operators can now model how power, cooling, and compute interact before a single piece of equipment is deployed. At the same time, the industry is entering what NVIDIA has called an “inference inflection point” as workloads move from training into large-scale, real-time deployment across industries—from industrial automation and robotics to geospatial intelligence and foundational digital infrastructure. Together, these shifts are redefining what AI infrastructure must deliver—and how precisely it must be designed in advance. But the transition from design to deployment remains the industry’s hardest problem.

This fireside brings together leaders across hyperscale infrastructure, AI systems design, and open hardware ecosystems to examine:

• How digital twins and simulation are shaping next-generation AI facilities

• How Meta and others are deploying AI infrastructure at global hyperscale

• The role of OCP and open systems collaboration in accelerating deployment

• Where designs break when they hit real-world construction, operations, and grid constraints

Framed through “designed vs. deployed,” the session looks at the widening gap—and emerging convergence—between what can be precisely modeled, and what can be reliably delivered at hundreds of megawatts.

Site selection for AI infrastructure is no longer an optimization problem with multiple good answers. It’s a constraint problem with no clean solutions. Power availability, fiber access, water resources, permitting timelines, and community acceptance are now tightly coupled variables—often in direct conflict. Every site presents tradeoffs, and the margin for error is shrinking.

At the same time, developers are beginning to invert the traditional model: siting infrastructure around energy availability itself and leveraging behind-the-meter strategies to open new regions.

This panel explores how site selection is actually happening today:

• Where projects are advancing—and why those locations clear the constraint stack

• Where projects are stalling on power, water, fiber, or community opposition

• How entitlement risk, zoning, and policy reshape “viable” vs. “theoretical” sites

• Which compromises developers are willing to accept—and which are fatal

Capital is pouring into AI infrastructure—but the assumptions behind that capital are being tested by execution reality.

Institutional investors are underwriting platform-scale deployments based on long-term AI demand. Yet delivering that capacity is proving far more complex. Tens of gigawatts have been planned for 2026 and beyond, but analysis suggests a material share will slip—not for lack of demand, but because the underlying infrastructure isn’t ready.

The constraint is no longer capital. It’s power—and the systems that deliver it. This session brings together perspectives across institutional capital allocation, development, energy-integrated infrastructure, and interconnection strategy to examine:

• How major platforms are underwriting power, supply chain, and timing risk

• Where projects are moving forward—and where they’re quietly being delayed or re-scoped

• Whether secondary and emerging markets represent durable opportunity or new uncertainty

• How network ecosystems and interconnection shape long-term value concentration

• How the rise of inference workloads may shift where and how capacity must be built

At stake is a key question: Is capital moving in step with what can be built—or ahead of the infrastructure required to support it?

Networking Lunch Break with Exhibitors

Power gets the headlines. Chips get the capital. But the AI buildout is increasingly gated by something less visible: the ability to move, stage, and deliver critical infrastructure at speed and at scale.

This session examines how global supply chains are becoming a defining constraint and differentiator in AI data center delivery. From long-lead electrical components to liquid cooling systems and mission-critical spares, execution now depends on tightly coordinated logistics across fragmented global networks.

Unilog brings a 4PL orchestration perspective on how leading operators are shifting from vendor management to system-level control. Additional panelists will provide viewpoints from development, operations, and infrastructure supply, creating a grounded discussion on what it actually takes to deliver in today’s constrained environment.

The AI factory is no longer a future concept. It is rapidly becoming the operating model for next-generation digital infrastructure.

But while the industry has become highly effective at designing AI environments on paper, the real challenge begins when high-density systems move into live production. Ultra-dense compute clusters, liquid cooling architectures, evolving power topologies, and massive interconnection demands are pushing infrastructure into operational territory few organizations have managed at scale before.

This panel brings together leaders across hyperscale operations, construction delivery, interconnection ecosystems, and AI systems architecture to examine what it actually takes to build, integrate, and operate AI infrastructure under real-world conditions.

Discussion topics include:

• What operational issues emerge first when AI workloads go live at scale

• How teams are adapting to liquid cooling and ultra-high-density deployments

• Where commissioning, integration, and operational handoffs create the greatest risk

• How networking, workload placement, and infrastructure orchestration are reshaping day-two operations

• Why operational execution is becoming a defining competitive advantage in AI infrastructure

As AI infrastructure evolves into industrial-scale production environments, the question is no longer simply who can design the AI factory — but who can successfully run it.

Networking Break with Exhibitors

The AI boom is forcing a fundamental rethink of how data centers acquire, distribute, cool, and manage power. What was once a utility procurement exercise is rapidly becoming a full-stack infrastructure challenge spanning generation, transmission, onsite power, power distribution, thermal management, and workload orchestration.

This session explores how operators are building the AI power stack from the grid connection all the way to the GPU. Panelists will examine the growing role of nuclear and gas generation, behind-the-meter power, battery storage, advanced electrical architectures, liquid cooling, and intelligent energy management systems.

The discussion will focus on what it takes to deliver reliable AI infrastructure at gigawatt scale while maintaining speed-to-market and operational resilience. Power availability, cooling efficiency, and integrated infrastructure design are increasingly becoming competitive advantages rather than engineering considerations.

Learning Objectives:

 - understand the evolving ai power stack

- evaluate emerging power generation technologies

- assess advanced energy infrastructure components

- examine integrated infrastructure design approaches

- identify key challenges in scaling to gigawatt-level infrastructure

Sponsor-led peer discussions designed for candid, high-value exchange. Each session features curated prompts in an interactive, no-slide format. Topics include:

• AI Data Centers: Density, Design, and the Reality of Deployment
• Data Center Construction: Speed, Scale, and Execution Under Pressure
• Liquid Cooling Challenges: From Early Adoption to Operational Reality
• Colocation in the AI Era: Capacity, Customers and the New Economics
• Onsite & Behind-the-Meter Power: Strategies for Speed, Certainty, and Control

Click here to register

For sponsorship opportunities, please contact Tom Larranaga, tlarranaga@endeavorb2b.com.

Continental Breakfast with Exhibitors

As AI deployments push rack densities beyond the limits of traditional data center designs, thermal infrastructure—including liquid cooling systems, mechanical plants, heat rejection technologies, water management strategies, and hybrid air-and-liquid environments—is rapidly becoming one of the industry's most critical capacity constraints.

The transition to AI-ready infrastructure extends far beyond the adoption of new cooling technologies. Operators must retrofit existing facilities, support increasingly diverse cooling architectures, manage water resources responsibly, prepare for higher-density workloads, and maintain operational resiliency as thermal demands continue to rise.

This session explores how developers, operators, and technology providers are scaling cooling infrastructure for the next generation of AI workloads. Panelists will discuss direct-to-chip cooling, facility retrofits, hybrid cooling environments, heat rejection strategies, water stewardship, operational readiness, and the lessons emerging from real-world AI deployments. From legacy data center upgrades to purpose-built AI campuses, the discussion will examine what it takes to build and operate the thermal infrastructure needed to support AI at scale.

As AI infrastructure scales, a new constraint is emerging: acceptance. In key markets, data center development is no longer judged on speed and capacity alone. Power consumption, land use, water, and grid impact are now front and center—shaping permitting timelines, community response, and long-term viability.

The challenge is not just opposition. It’s alignment—and credibility. Communities are responding less to what the industry says it will do, and more to what it actually delivers: how projects are designed and built, how resources are used, how transparently developers engage, and whether local economic benefits are real and durable.

This closing session reframes “social license” as an outcome of execution, not messaging.

Topics include:

• Where and why data center projects are facing pushback in today’s market • How early decisions on siting, power, and design shape community outcomes years later

• What responsible growth looks like under real power, land, and policy constraints

• How workforce development, local participation, and construction practices influence trust

• Whether current frameworks and narratives are keeping pace with the scale and visibility of AI-led development

The discussion focuses on what actually builds trust—and what erodes it—recognizing that the next phase of growth won’t be decided by demand alone, but by whether communities will accept the infrastructure required to support it.

We are NTT Global Data Centers. As the third largest data center provider, we operate over 160 data centers in more than 20 countries and regions. We understand that every business – large and small – has its own unique needs and goals. We offer local-to-global data center expertise, aligned with our connected platform of AI-ready data centers to create solutions that enable our clients to seamlessly scale their digital businesses, anywhere and anytime.

Note: Photography is not allowed except in approved areas TBD by NTT.

Cost: $100 per person

Ashburn Data Center is a 38-acre data center campus residing in the heart of the nation’s densest connectivity corridor. Located in Ashburn, VA in Loudoun County (commonly referred to as Data Center Alley), the campus is less than an hour drive to Washington, D.C. Two of three buildings are now completed and leasing collocation and powered shell data center space, offering tenants over 36 MW of power and access to multiple connectivity options.

Cost: $100 per person