Cooling the 1-2 MW Rack: Liquid Strategies That Actually Deploy




Already registered? Click here to log in.
Summary
AI infrastructure is rapidly pushing beyond the limits of traditional air-cooled data center design. With rack densities now approaching 1 MW, liquid cooling is no longer experimental: it is becoming a core requirement for deploying next-generation AI clusters at scale.
Yet many operators, developers, and enterprise buyers still face practical questions: What designs actually deploy? What scales reliably? And what can be supported and serviced over time?
This Data Center Frontier webinar moves beyond theory to focus on real-world liquid cooling strategies now entering production environments. Industry experts will discuss cooling architectures, operational tradeoffs, supply chain readiness, and service models that enable high-density AI deployments to move from pilot projects into repeatable infrastructure.
Key discussion areas include:
• Direct-to-chip vs. immersion: what operators are deploying today
• CDU and facility loop design for megawatt-class racks
• Warm-water cooling and facility integration strategies
• Serviceability, maintenance, and operational risk management
• Supply chain readiness for pumps, manifolds, fluids, and controls
• Brownfield retrofits vs. greenfield AI campuses
• Designing infrastructure that survives multiple hardware generations
As AI campuses scale from megawatts per hall to hundreds of megawatts per site, cooling design is becoming a central business and operational decision. This session provides practical guidance on deploying liquid cooling solutions that work today, and scale for tomorrow’s density requirements.
Who should attend: Data center operators, hyperscale infrastructure teams, colocation providers, developers, enterprise infrastructure leaders, design engineers, investors, and vendors building infrastructure for the AI era.









