

The state of play of data center development
34 snips Jun 12, 2025
In this engaging discussion, Chris Sharp, Chief Technology Officer of Digital Realty, shares insights on the evolving landscape of data centers. He delves into how AI's demands create 'data gravity' and influence infrastructure development. The conversation covers the balancing act between large builds and incremental improvements, the reality behind power needs versus hype, and the role of 'bridge power' in today's energy landscape. Sharp emphasizes the critical interplay between data center innovation and the future grid, addressing emerging challenges in a rapidly changing tech environment.
AI Snips
Chapters
Transcript
Episode notes
Training vs. Inference Data Center Needs
- AI workloads bifurcate into training and inference, affecting data center location choices.
- Training is flexible with location; inference demands proximity for latency and throughput.
Data Gravity Shapes AI Centers
- Data gravity and data oceans attract compute infrastructure, concentrating AI workloads.
- AI inference is clustering near existing data centers and clouds due to ecosystem needs.
Power Limits Data Center Growth
- Power availability is often the first constraint in major data center markets like Northern Virginia.
- Utility grid distribution, not just generation, limits capacity requiring new solutions.