Idem Est Research Launches AI Demand & Connectivity Atlas, Forecasting Australia's Data Centre IT Load to Reach 10.5 GW by 2035

Sydney, Australia, 11 May 2026 - Idem Est Research & Advisory today announced the launch of the AI Demand & Connectivity Atlas, a strategic market intelligence platform designed to forecast where AI-driven data centre demand will land across Australia and Asia-Pacific.

The Atlas links data centre supply, workload demand, gigawatt-scale growth and network intensity to model how AI training, AI inference, cloud, video and physical AI workloads are expected to reshape APAC data centre clusters. The platform is designed for fibre and network providers, investors, infrastructure strategy teams, public sector agencies and advisory firms seeking a forward-looking view of AI infrastructure demand.

Australia is emerging as one of APAC's leading AI infrastructure markets, with its combination of hyperscale cloud expansion, deep data centre pipelines, renewable energy potential, major metro clusters and growing demand for low-latency connectivity. The Atlas forecasts Australia's IT load rising from 1.5 GW in 2025 to 10.5 GW by 2035, representing an estimated 22% compound annual growth rate, with AI-related workloads accounting for approximately 55% of modelled 2035 IT load.

The Australian DC Database identifies 166 operational data centre sites and 106 pipeline sites, with Sydney as the largest current hub, Melbourne as the secondary hub, Brisbane as an emerging hub and Canberra as a secure and government workload node.

"AI infrastructure growth is no longer only a data centre capacity question. It is becoming a power, fibre, backhaul, interconnect and regional planning question," said Landry Fevre, Managing Director, Idem Est Research & Advisory. "Australia is already one of the region's most important AI infrastructure markets, and the Atlas shows how AI demand is translating into gigawatt-scale IT load growth and significant connectivity pressure across key metro clusters."

Australia's AI infrastructure market outlook

Australia's forecast highlights the scale of infrastructure planning required as AI workloads increase their share of national data centre demand.

The Atlas models Australia's IT load and capacity baseline at 1.5 GW in 2025, rising to a forecast 10.5 GW by 2035. This represents an estimated 22% compound annual growth rate over the period.

Idem Est Research's Australian DC Database identifies 166 operational data centre sites and 106 pipeline data centre sites across the country.

Sydney remains Australia's largest current data centre hub, while Melbourne is the country's secondary hub. Brisbane is identified as an emerging hub, and Canberra is identified as a secure and government workload node.

Forecast outputs are designed to support strategy, prioritisation and planning. They should be used alongside site-specific engineering, power-procurement and network-design studies.

AI workloads forecast to drive 55% of Australia's 2035 modelled IT load

The Atlas forecasts AI-related workloads becoming the largest driver of future data centre demand in Australia.

By 2035, AI inference is forecast to account for 36% of modelled IT load, making it the largest individual workload category. AI training is forecast to represent 14%, while physical AI and edge workloads are forecast to represent 5%. Together, these AI-related categories account for approximately 55% of modelled 2035 IT load.

Other major workload categories include general cloud at 26%, CDN and video workloads at 17%, and enterprise / WAN workloads at 3%.

The Atlas models AI training as more likely to concentrate in larger campus-scale clusters, while AI inference demand is expected to emerge more broadly across metro nodes requiring lower-latency connectivity.

This shift has direct implications for data centre site selection, grid planning, metro fibre deployment, interconnection strategy and long-haul transport network investment.

Sydney West and Melbourne West emerge as major connectivity pressure points

The Atlas' connectivity layer estimates where AI, cloud and content workloads are likely to create the greatest network pressure.

In Australia, Sydney West is ranked as the leading connectivity hotspot, with modelled aggregate connectivity intensity of 1,340 Tbps under the selected workload-intensity scenario.

Melbourne West ranks second, with modelled connectivity intensity of 770 Tbps.

Sydney West / Aerotropolis ranks third, with modelled connectivity intensity of 493 Tbps.

For network providers, the connectivity layer supports fibre route prioritisation, metro backhaul planning, interconnect strategy and peering ecosystem analysis. The Tbps figures represent modelled aggregate connectivity demand by cluster under the selected workload-intensity scenario.

Beyond site listings: modelling where demand is going

While many data centre datasets focus on where facilities are located, the AI Demand & Connectivity Atlas is designed to show where demand is expected to concentrate. It combines a bottom-up data centre market view with workload allocation, utilisation and ramp-up assumptions to estimate total IT load, AI share of demand, workload mix and connectivity implications by cluster and corridor.

The platform provides three linked market views.

The first is a workload-based MW forecast, providing a forward view of data centre demand by cluster, state and region.

The second is an AI workload mix, showing demand split across AI training, AI inference, physical AI, cloud, enterprise, video and content workloads.

The third is a connectivity demand outlook, estimating Tbps implications derived from power and workload intensity to support corridor and cluster prioritisation.

Built for infrastructure strategy, investment and planning

The AI Demand & Connectivity Atlas has been developed to help infrastructure stakeholders identify where AI load is concentrating fastest, compare clusters by future digital infrastructure demand, prioritise markets for fibre, backhaul and interconnect investment, challenge-test public demand assumptions with a bottom-up market benchmark, and support executive strategy, investment planning and infrastructure prioritisation.

Best-fit users include fibre and network infrastructure providers, investors, digital infrastructure strategy teams, government and infrastructure agencies, advisory firms and research teams.

Available across APAC with custom market scopes

The same modelling approach can be applied across selected APAC markets, including Australia, New Zealand, Singapore, Malaysia, Indonesia, the Philippines, Thailand and Vietnam. Outputs can be scoped at national, state, metro, corridor or custom-cluster level.

The Atlas is delivered on a proposal-led and custom-scoped basis, with optional quarterly updates and delivery formats including executive briefings, CSV, KML/KMZ, GeoJSON and shapefile outputs.

More information is available on the AI Demand & Connectivity Atlas page.
https://idemest.com/reports/ai-demand-connectivity-atlas/

Landry Fevre
Idem Est Research & Advisory
Woonona Rd Northbridge, NSW 2063, Australia
Email: landry@idemest.com
Phone: +61 414 164 100

About Idem Est Research & Advisory
Idem Est Research & Advisory is a boutique TMT research and advisory firm with 25+ years' experience specialising in Asia Pacific telecommunications and digital infrastructure research

This release was published on openPR.