Pre-Conference Day

• Evaluating air, liquid, and hybrid cooling architectures alongside CDU configurations to align with AI chip demands, cabinet design, and evolving deployment models
• Assessing Eastern State climate impacts, heat rejection limits, and chiller derating to mitigate thermal risk and maintain performance during peak AI load conditions
• Integrating thermal storage, buffer tanks, and mature piping, and vendor strategies to stabilize system hydraulics and reduce operational risk as cooling technologies evolve

• Demonstrating how major scopes of work were shifted from site to controlled manufacturing environments to deliver more capacity with fewer on-site labour resources
• Quantifying where modularization materially reduced program risk and where it did not, based on what was implemented rather than discussed in theory
• Translating owner, contractor, and design-side decisions into repeatable approaches that show what “doing more with less people” looks like in practice

As AI workloads surge ahead of grid capacity, this session brings the industry face-to-face with the realities of constrained power. In a fast-moving, interactive roundtable format, you’ll dig into the pressure points – from interconnection delays to policy headwinds – and explore where smarter redundancy, architecture, and cooling strategies can unlock deployment without sacrificing uptime. Expect candid discussion on the trade-offs shaping site design today, and practical insight into how to build infrastructure that performs under pressure now while staying flexible as AI demands keep evolving.