By Mark Rafferty, CEO, FibreconX
When we were designing FibreconX, the conventional wisdom in the market was straightforward: build data centres, and the network will follow. Connectivity was always downstream – a commodity you’d procure once the real infrastructure decisions were made.
We made a different bet. We believed demand would outrun infrastructure and we built accordingly.
Watching what’s happening in the Australian market right now, I’m more convinced than ever that the supply-demand mismatch in network infrastructure is the story that isn’t being told loudly enough.
The Numbers Tell a Stark Story
Australia is in the middle of a data centre investment surge unlike anything the market has seen. Billions in capital commitments, new campuses announced across Melbourne and Sydney and capacity scaling rapidly across both cities.
But beneath that headline growth is a structural issue that deserves more attention. According to the Grandview Research Australia Dark Fibre Networks Market Report, the Australian dark fibre market is forecast to grow at 17% per year through to 2030 and demand in the data centre interconnect segment is accelerating even faster than that broader trend with AI workloads accounting for approximately 50% of the total demand growth.
That’s not linear growth. It’s compounding demand across both compute and connectivity simultaneously, in the same corridors, competing for the same constrained physical infrastructure.
Where the Model Breaks
The assumption that connectivity follows compute was reasonable in a previous era. In that world, you’d build a data centre, run some conduit, light some fibre, and the network was done. Connectivity was a feature of the facility, not a gating factor on it.
That model no longer holds. Legacy fibre infrastructure in Australia’s key metro corridors is already heavily utilised. Available spare capacity is limited and critically, much of what does exist is locked behind lit service models — managed, shared, inflexible — rather than scalable dark fibre that operators can dimension themselves.
The result is a classic supply-demand imbalance. Demand for high-capacity, flexible, scalable fibre is accelerating. Supply of available infrastructure that can actually meet those requirements is constrained, and the gap is widening.
“Demand would always outrun infrastructure. The winners would be the ones who built before the crowd arrived.”
A Supply Constraint Hiding in Plain Sight
For operators, wholesalers, and hyperscalers, this constraint has real commercial consequences. Slower deployment timelines when you’re trying to connect facilities. Higher costs when you’re competing for limited available capacity. And limited flexibility when your business needs to scale faster than a legacy provider’s upgrade cycle allows.
Site selection decisions are increasingly being made based on connectivity availability. Speed to market in the data centre business now depends as much on network readiness as it does on power and cooling. Long-term scalability requires infrastructure that can grow with you, not infrastructure that needs to be replaced when you outgrow it.
Pre-Building for Demand, Not Reacting to It
The decision we made at FibreconX was to build ahead of demand, significantly ahead. That required capital discipline and conviction in our market thesis. It also required building differently: not just more fibre, but the right kind, in the right places, through infrastructure we own and control.
In Sydney and Melbourne alone, we’ve built independent duct infrastructure across the key metro corridors where data centre demand is concentrating. And while legacy networks typically deploy just 72, 144, or even 288 fibres per data centre, we’ve taken a fundamentally different approach, averaging 14,000 fibres to every data centre, delivered across an average of 4 fully diverse paths within our own dedicated duct network. Current utilisation sits well below 5%, and that’s entirely by design. It’s not a sign that demand isn’t there – it’s the headroom we deliberately built to serve the demand that’s coming.
The organisations that treat network availability as a primary infrastructure consideration, not an afterthought, will select better sites, move faster, and scale more efficiently than those who discover the constraint after the lease is signed.
→ Explore FibreconX network availability in Sydney and Melbourne