Data Center Priority 3: AI Infrastructure (WWT Research)

3,000 Pounds of Silicon: The Physical Limits of AI Infrastructure

AI hardware doesn't care about your quarterly goals. It cares about physics.
More than 90% of enterprise data centers can't run next-generation AI hardware right now. Not in six months — today. Which means most organizations planning to deploy AI workloads on-premises are working with assumptions that don't hold up to the laws of physics.
In this episode, Chris Campbell (Senior Director of AI Solutions at WWT) walks through what happens when the urgency to secure cutting-edge chips outruns the ability to actually power them. We're talking about Blackwell pods sitting idle on loading docks, raised floors collapsing under 3,000-pound supercomputers, and 12-year timelines just to complete electrical grid interconnect studies.

What you'll learn:
- Why retrofitting existing data centers for AI workloads is often a non-starter
- The real economics of cloud vs. neo-cloud vs. on-prem AI infrastructure
- How power availability is becoming a strategic advantage (and why Australia matters)
- The six-step framework for AI-ready infrastructure planning
- Why your most important strategic partner might be a power utility in Oklahoma

Key takeaway: It takes three years to build a new data center, but only three minutes to order hardware you can't plug in. The interim strategy isn't just a bridge — it's your survival plan for the next thousand days.

Featured Guest:
Chris Campbell, Senior Director of AI Solutions, World Wide Technology

Resources:
📄 Download Priority 3: AI Readiness research paper → https://www.wwt.com/wwt-research/it-infrastructure-modernization-priorities-for-2026
🔬 Explore WWT's $500M AI Proving Ground with 17+ reference architectures

Timestamps:
0:00 – The physics problem
2:15 – The Blackwell dock scenario
5:42 – Why 90% of data centers aren't ready
8:20 – Retrofitting vs. building new
12:35 – The power constraint reality
16:10 – Global infrastructure and data sovereignty
19:45 – Six practical steps (and the readiness assessment)
24:30 – Skills gap and workforce pivot opportunities

This episode is part of WWT's Data Center Priorities research series. Episode supported by Graphient.
#AIInfrastructure #DataCenter #EnterpriseAI #WWTResearch #AIReadiness

Data Center Priority 3: Artificial Intelligence
WWT 2603

Видео Data Center Priority 3: AI Infrastructure (WWT Research) канала ExplaiNerds