LATEST NEWS

DataBank and Goodman Group Partner to Open Los Angeles Data Center. Read the press release.

Hinder the Future of AI
Hinder the Future of AI

Hinder the Future of AI

  • Updated on October 12, 2023
  • /
  • 1 min read

Built In explores how data center shortages could constrain AI development, with DataBank CEO Raul Martynek examining four critical bottlenecks: capital costs, construction time, space availability, and power constraints. The analysis reveals data center vacancy rates dropped to 2.88 percent in North America’s top markets as generative AI demands five times the power of traditional workloads.

The article details how a single medium-sized data center can cost nearly half a billion dollars to construct, requiring 24-to-36 months to complete amid supply chain challenges. GPU-powered HPC clusters for AI training are competing with traditional enterprise and hyperscale cloud growth for limited capacity.

“What happens when there is more demand for AI than what our current infrastructure can support? The answer seems simple: overcome a lack of capacity by building more capacity. In practice, however, it becomes much more difficult.”
— Raul Martynek, CEO of DataBank

The shortage could hamper AI application deployment, strain existing cloud infrastructure, and challenge companies like Nvidia despite unprecedented GPU demand. Smart data center providers are responding by increasing efficiency and positioning facilities near power generation sources.

Explore the complete infrastructure analysis at Built In.

Share Article


Get Started

Discover the DataBank Difference today:
Hybrid infrastructure solutions with boundless edge reach and a human touch.