Panel: From compute clusters to global AI corridors

AI infrastructure is currently planned and delivered across separate industries. Compute, data centres, power, subsea systems, terrestrial fibre and cloud interconnection often have different investment cycles, commercial models and deployment timelines. This session explores how the ecosystem can plan the complete corridor rather than its individual assets. 

• What are the essential components of an AI-ready international corridor?  

• Where are the weakest links in today’s infrastructure?  

• How should network deployment align with hyperscaler and neocloud compute roadmaps?  

• What role should data-centre operators play in corridor planning?  

• Where do metro infrastructure and cloud interconnection create hidden bottlenecks?  

• How important are subsea and landing-station diversity to AI infrastructure decisions?  

• Who should coordinate the ecosystem when no single company controls the entire corridor?  

• How can infrastructure partners share demand signals without compromising commercial confidentiality?