Activation status
Please contact Vertiv Partner Support: salescloudsupport@vertivco.com
Working with our Vertiv Sales team enables complex designs to be configured to your unique needs. If you are an organization seeking technical guidance on a large project, Vertiv can provide the support you require.
Many customers work with a Vertiv reseller partner to buy Vertiv products for their IT applications. Partners have extensive training and experience, and are uniquely positioned to specify, sell and support entire IT and infrastructure solutions with Vertiv products.
Already know what you need? Want the convenience of online purchase and shipping? Certain categories of Vertiv products can be purchased through an online reseller.
Need help choosing a product? Speak with a highly qualified Vertiv Specialist who will help guide you to the solution that is right for you.
-
Home
- About
- News & Insights
- Events
- DCD Power Innovation Day 2026
DCD Power Innovation Day 2026
Power strategies for the next generation of AI infrastructure
AI infrastructure is changing how data centers distribute, manage, and plan for power. Increasing rack densities, dynamic AI load profiles, and evolving processor architectures are introducing new considerations across compute, facility power, cooling, and utility integration.
Join Vertiv during DCD Power Innovation Week for a three-part discussion examining approaches for maximizing AI output from available power resources, evaluating emerging power architectures, and coordinating AI infrastructure with the broader power ecosystem.
Why attend?
Participants will gain insights into:
- Measuring AI infrastructure performance through productive AI output
- Maximizing compute and sellable capacity from each provisioned megawatt
- Evaluating the role of 800 VDC architectures in future AI power strategies
- Planning for increasing rack densities and evolving GPU requirements
- Managing dynamic AI load profiles
- Coordinating utility power, energy storage, onsite generation, cooling, and AI workloads
- Assessing AI power smoothing, load flexibility, and microgrid strategies
- Developing a systems-level approach to AI infrastructure
Featured episodes
Peter Panfil
Vice President, Technical Business Development, Vertiv
Phillip Marangella
Chief Marketing and Product Officer, EdgeConneX
Martin Olsen
Vice President, Segment Strategy and Deployment, Vertiv
Giuseppe Leto
Senior Director, IT Systems Business - EMEA
Tokenomics: Maximizing AI output per megawatt
Tokenomics reframes AI infrastructure performance around productive AI output, not just dollars per watt or facility efficiency. One emerging metric is tokens per sec per MW, which connects useful inference throughput to provisioned utility power and shifts the conversation toward maximizing AI output on the lowest-TCO compute infrastructure. This discussion explores how two levers, improving peak PUE to increase sellable megawatts, and improving IT efficiency to increase compute per megawatt, can influence AI factory performance. We will cover:
- The practical framework for understanding how power, cooling, processor architectures, and IT optimization work together
- How to maximize the AI value generated from every contracted megawatt
- How to optimize tokens per second per MW
- Real-world strategies for reducing TCO while increasing inference throughput
Ke Sun
Senior Technical Business Development Manager, Vertiv
Tam Pledger
Digital Content Manager, DCD
Sanjay Sainani
Senior Vice President, Technology and Data Centers Business Development, Vertiv
Giuseppe Leto
Senior Director, IT Systems Business - EMEA
The 800 VDC decision for AI power architecture
GPU rack density is rising faster than infrastructure planning cycles. Today's AI training racks draw 145 kW, with next-generation platforms expected to reach 600 kW, and future platforms exceeding 1 MW. In this session, we answer the key questions that every operator planning AI infrastructure needs to address to prepare their power architecture. The session will focus on open, practical design criteria and will cover:
- Who needs 800 VDC today, and who does not
- Why architecture optionality matters more than picking one path early, and which market claims deserve scrutiny
- How the phased journey to 800 VDC power architecture works
- Key considerations beyond power efficiency
AI rack densities continue to increase, creating new power delivery considerations. As infrastructure requirements evolve, operators are evaluating how power architectures can address higher-density deployments and future scalability requirements.
This episode examines practical considerations for organizations evaluating 800 VDC architectures. The discussion will address where 800 VDC architectures may be applicable, where AC-based architectures remain a practical option, why architecture optionality matters, and how power architectures may evolve through sidecar, pod-level, and potential centralized DC implementations.
Chris Thompson
Vice President, Global Advanced Technology and Microgrid Solutions, Vertiv
Giovanni Zanei
VP, Large Power Conversion, Vertiv
Giovanni Zanei
Vice President, Large Power Conversion, Vertiv
Giuseppe Leto
Senior Director, IT Systems Business - EMEA
Grid-Interactive AI: Rethinking the relationship between data centers and the grid
AI data centers are no longer passive loads. Dynamic AI load profiles are changing how facilities interact with utilities, onsite energy resources, and critical infrastructure. This episode will explore the spectrum of grid-interactive strategies available to AI operators, from AI power smoothing and fault ride-through to energy storage, load flexibility, and microgrid orchestration. Join us to unpack:
- The practical framework for understanding how coordinated power management can improve resilience and support utility relationships
- How to accelerate interconnection and time-to-power
- How to maximize the value of available energy resources across a range of deployment scenarios
- Coordinating AI workloads, energy storage, and onsite generation to balance flexibility, reliability, and cost
Peter Panfil
Vice President, Technical Business Development, Vertiv
Martin Olsen
Vice President, Segment Strategy and Deployment, Vertiv
Phillip Marangella
Chief Marketing and Product Officer, EdgeConneX
Tokenomics: Maximizing AI output per megawatt
Tokenomics reframes AI infrastructure performance around productive AI output, not just dollars per watt or facility efficiency. One emerging metric is tokens per sec per MW, which connects useful inference throughput to provisioned utility power and shifts the conversation toward maximizing AI output on the lowest-TCO compute infrastructure. This discussion explores how two levers, improving peak PUE to increase sellable megawatts, and improving IT efficiency to increase compute per megawatt, can influence AI factory performance. We will cover:
- The practical framework for understanding how power, cooling, processor architectures, and IT optimization work together
- How to maximize the AI value generated from every contracted megawatt
- How to optimize tokens per second per MW
- Real-world strategies for reducing TCO while increasing inference throughput
Ke Sun
Senior Technical Business Development Manager, Vertiv
Tam Pledger
Digital Content Manager, DCD
Sanjay Sainani
Senior Vice President, Technology and Data Centers Business Development, Vertiv
The 800 VDC decision for AI power architecture
GPU rack density is rising faster than infrastructure planning cycles. Today's AI training racks draw 145 kW, with next-generation platforms expected to reach 600 kW, and future platforms exceeding 1 MW. In this session, we answer the key questions that every operator planning AI infrastructure needs to address to prepare their power architecture. The session will focus on open, practical design criteria and will cover:
- Who needs 800 VDC today, and who does not
- Why architecture optionality matters more than picking one path early, and which market claims deserve scrutiny
- How the phased journey to 800 VDC power architecture works
- Key considerations beyond power efficiency
AI rack densities continue to increase, creating new power delivery considerations. As infrastructure requirements evolve, operators are evaluating how power architectures can address higher-density deployments and future scalability requirements.
This episode examines practical considerations for organizations evaluating 800 VDC architectures. The discussion will address where 800 VDC architectures may be applicable, where AC-based architectures remain a practical option, why architecture optionality matters, and how power architectures may evolve through sidecar, pod-level, and potential centralized DC implementations.
Chris Thompson
Vice President, Global Advanced Technology and Microgrid Solutions, Vertiv
Francesco Mangano
Head of Engineering, Data4
Giovanni Zanei
Vice President, Large Power Conversion, Vertiv
Grid-Interactive AI: Rethinking the relationship between data centers and the grid
AI data centers are no longer passive loads. Dynamic AI load profiles are changing how facilities interact with utilities, onsite energy resources, and critical infrastructure. This episode will explore the spectrum of grid-interactive strategies available to AI operators, from AI power smoothing and fault ride-through to energy storage, load flexibility, and microgrid orchestration. Join us to unpack:
- The practical framework for understanding how coordinated power management can improve resilience and support utility relationships
- How to accelerate interconnection and time-to-power
- How to maximize the value of available energy resources across a range of deployment scenarios
- Coordinating AI workloads, energy storage, and onsite generation to balance flexibility, reliability, and cost
Explore power innovation for AI infrastructure
Join Vertiv during DCD Power Innovation Week on October 16, 2026, for discussions on AI infrastructure performance, emerging power architectures, and the evolving relationship between data centers and the grid.
Learn how a systems-level approach spanning compute, power, cooling, energy storage, onsite generation, and utility integration can help operators address changing AI infrastructure requirements.
Register to stream