Vertiv leaders join NVIDIA ecosystem partners, infrastructure experts, and industry researchers across five DCD webcasts examining how operators can turn AI ambition into deployable, productive, and adaptable capacity.
AI infrastructure has moved past the question of whether demand is real. The question now is how to meet it: where workloads land, how much power operators can secure, what densities facilities can support, and how to avoid stranded capital. These are not abstract concerns. They are the decisions operators across hyperscalers, colocations, NeoClouds, and enterprises are negotiating right now.
That is the backdrop for DCD AI Week. The series runs from Monday, July 27, to Friday, July 31, on the DCD platform, with one webcast each day. Together, the five sessions follow the AI infrastructure decision journey: understanding the market shift, securing and orchestrating power, validating the design before construction, assessing how enterprise workloads drive deployment decisions, and preserving performance and optionality throughout the operating lifecycle. Here is what to expect.
Day one: From planning and design to deployment reality
The series opens Monday, July 27 with From planning and design to deployment reality — What operators must solve next, streaming from 09:00–10:00 EDT on the DCD Compute, Storage & Networking Channel. Vertiv CEO Giordano Albertazzi joins DCD CEO Dan Loosemore to discuss how AI infrastructure is moving from experimentation to deployment at scale.
The session focuses on the pressures operators face now: density, speed, scale, complexity, and dynamic load profiles. It also explores the shift from measuring announced or installed megawatts to maximizing useful AI output from every available megawatt, with tokens per second per megawatt as one supporting metric; how requirements differ across hyperscalers, NeoClouds, colos, and enterprises; and why the market is moving from product selection to validated system architectures.
Day two: Delivering power confidence for high-density data centers
Tuesday, July 28, turns to one of the most significant constraints in AI infrastructure today: power. The webcast, Delivering power confidence for high-density data centers, runs from 12:00–13:00 EDT on the DCD platform. George Hannah, Vice President of Thermal Systems and Advanced Development at Vertiv, and Chris Thompson, Vice President, Global Advanced Technology and Microgrid Solutions, Vertiv, join Arnaud Van Der Haegen, Manager of Power Generation at Solar Turbines, in a discussion moderated by James Raddings, Digital Portfolio Lead at DCD.
AI workloads create dynamic power demands that challenge conventional assumptions about generation, protection, storage, and cooling. The session will examine how operators can combine utility power, on-site generation, battery energy storage systems, UPS, evolving voltage architectures, and controls across different response times to bring capacity online in phases and reduce deployment risk. The central message: build a coordinated power strategy, not a collection of backup assets.
Day three: From blueprint to build — designing smarter with reference architectures and digital twins
On Wednesday, July 29, the conversation moves to design methodology. The webcast, From blueprint to build — Designing smarter with reference architectures and digital twins, runs from 11:00–12:00 EDT on the DCD Compute, Storage & Networking Channel. Viktor Petik, Senior Vice President of Infrastructure Solutions at Vertiv, joins Stephane Sireau, VP, High Tech Industry at Dassault Systèmes, in a discussion moderated by Emma Strutton, Head of Channels at DatacenterDynamics.
This session examines how validated reference architectures, model-based systems engineering, simulation-ready assets, and virtual twins can identify interface risks before they become construction or commissioning problems. It will explore how compute blueprints such as NVIDIA DSX can anchor power, thermal, controls, and physical-layout decisions, and how digital continuity can preserve design intent from initial planning through operation.
Day four: Enterprise AI ‒ from pilot to production
Thursday, July 30, focuses on moving enterprise AI from pilots and productivity tools into production. The webcast, From pilot to production — Understanding how enterprise AI workloads drive infrastructure decisions, runs from 10:00–11:00 EDT on the DCD Compute, Storage & Networking Channel. Martin Olsen, VP of Segment Strategy and Deployment at Vertiv, joins Diana Kearns-Manolatos, Associate Director at Deloitte, and Matt Jacobs, Managing Director at Deloitte, in a discussion moderated by Zoe Turner, Channel Manager at DatacenterDynamics.
As enterprise AI moves into production, infrastructure choices increasingly depend on workload behavior, data gravity, latency, governance, utilization, cost predictability, and business criticality. The session will examine how those factors determine whether workloads run in public cloud, colocation, enterprise facilities, at the edge, or across a hybrid model, and how rack, power, cooling, and whitespace constraints shape those decisions.
Day five: The AI data center operating model — performance, flexibility, and future readiness
The series closes on Friday, July 31, with The AI data center operating model — performance, flexibility, and future readiness, streaming from 09:00–10:00 EDT on the DCD Compute, Storage & Networking Channel. Scott Armul, Chief Product & Technology Officer at Vertiv, joins Jim McGregor, Founder and Principal Analyst at TIRIAS Research, in a discussion moderated by Emma Strutton, Head of Channels at DatacenterDynamics.
The closing session examines how operators can maintain performance while compute, power, cooling, controls, and service requirements evolve across successive technology generations. The discussion will focus on lifecycle assurance, controls complexity, skills, predictive service, upgrade planning, and how to preserve architectural optionality without overbuilding today.
Why this week matters
AI infrastructure decisions made today will determine how much capacity becomes productive, how quickly it comes online, and how readily it can adapt to the next compute generation. Across five sessions, DCD AI Week will examine the decisions operators must connect rather than make in isolation: workload placement, power strategy, system validation, physical infrastructure design, and lifecycle operation.
Register now for the DCD AI Week sessions. All five DCD webcasts are available from July 27 to 31.