The AI in Business Podcast Scaling GenAI Without Burdening the Data Center - with Greg Ratcliff of Vertiv
26 snips
Nov 6, 2025 In this discussion, Greg Ratcliff, Chief Innovation Officer at Vertiv with three decades of experience in data and IoT, shares insights on scaling AI through smart infrastructure. He highlights the significance of addressing power, latency, and throughput constraints. Greg advocates for modular, prefabricated data centers to tackle talent shortages. He also emphasizes the value of hybrid data strategies and the importance of keeping sensitive data local. Misconceptions about data center careers are addressed, portraying them as high-tech and sustainable job opportunities.
AI Snips
Chapters
Transcript
Episode notes
Infrastructure Often Limits AI Progress
- Throughput, latency, and power constraints often block AI deployments more than algorithmic limits.
- Organizations must shift focus from models to the infrastructure that enables them.
Buy Prefabricated Infrastructure Modules
- Invest in modular designs and prefabricated systems to reduce on-site labor and deployment time.
- Use skidded, pre-manufactured components so teams can truck ready systems to sites quickly.
Design Talent Exists; Deploy Talent Lags
- There's a skills pipeline for design but a shortage of deployment-level technicians like electricians and network installers.
- The industry needs to improve reputation and outreach to attract those practical talent pools.