Understand how AI is reshaping data centre infrastructure and what it takes to build for the next wave of growth
AI demand is increasing, but compute is only part of the challenge. In this episode of Uplink, Michael Reid speaks with Tony Rossabi, Founder and CEO of OCOLO, about the infrastructure constraints shaping the next wave of growth, including power, data centre capacity, cooling, fibre connectivity, and deployment speed. Learn how infrastructure teams are adapting to support AI, cloud, and modern digital workloads at scale.
Key takeaways
As AI adoption accelerates, the conversation around infrastructure is shifting beyond GPUs and compute. Michael Reid speaks with Tony Rossabi, Founder and CEO of OCOLO, about the practical challenges facing data centre operators as demand for AI, cloud, and digital services continues to grow.
The discussion explores how power, cooling, fibre connectivity, and deployment timelines are becoming critical constraints, while also examining how infrastructure providers are preparing for the next generation of AI-ready facilities.
Why is AI changing data centre design?
AI workloads require significantly higher rack densities, more power, advanced cooling technologies, and faster deployment than traditional enterprise infrastructure. The discussion explains why infrastructure planning now extends far beyond compute capacity and why operators are redesigning facilities to support modern AI and cloud workloads.
Power, cooling and connectivity are becoming strategic infrastructure
The conversation highlights that access to power, fibre connectivity, land, and cooling systems is increasingly determining how quickly organisations can deploy new capacity. These infrastructure layers are becoming just as important as compute itself when planning for long-term growth.
Is AI the only driver of infrastructure investment?
No. While AI is accelerating demand, Tony explains that cloud adoption, enterprise modernisation, streaming services, and digital transformation continue to create sustained demand for data centre infrastructure. Understanding these broader market drivers helps organisations make more balanced investment decisions.
Looking ahead: What will future infrastructure look like?
As AI inference becomes more distributed, infrastructure will need to evolve alongside it. Compute, Network, and Storage must work together to deliver resilient, software-provisioned infrastructure capable of supporting global digital workloads.
