CoreWeave Targets 360MW Of AI Cloud Capacity In Indonesia By 2028
Data Center Knowledge reported that CoreWeave is using Indonesia as its first Asia-Pacific data centre market, tying the move to regional AI demand, latency needs and data-control requirements.
CoreWeave plans to make Indonesia its first Asia-Pacific data centre market, with three Greater Jakarta facilities totaling 360 MW of contracted IT power expected to begin service in 2028, Data Center Knowledge reported.
The move takes the GPU-cloud company beyond its North American and European footprint at a time when AI infrastructure demand is becoming more regional.
The Jakarta sites are aimed at a Southeast Asian customer base that includes AI-native builders, companies and public-sector buyers.
Jakarta Becomes The Entry Point
The three sites will be leased from a colocation provider whose name was not disclosed.
A company spokesperson put the combined critical IT load at 360 MW, with CoreWeave retaining ownership and operational control of the compute layer.
The deployment adds an Asia-Pacific node to a platform that had 49 data centres worldwide as of March 31, 2026.
CoreWeave listed active power above 1 GW and contracted power above 3.5 GW for AI workloads at that point.
Indonesia was chosen over more established regional hubs because of its AI ecosystem, digital infrastructure investment and education push.
The plan also includes hiring and training an operations team in the country.
Regional AI Demand Changes The Footprint
The workload mix points beyond massive centralized training capacity toward inference, lower-latency services and closer data control.
Data residency rules, inference demand and government AI projects are pushing cloud infrastructure providers to put compute nearer to users and data.
IDC cloud and edge infrastructure research vice president Dave McCarthy described the Indonesia plan as a global-platform signal.
CoreWeave is moving from a North American and European position into competition with legacy hyperscalers in more regions, with the APAC scale giving it an early position as enterprise AI adoption builds toward 2028.
Indonesia also offers land, power availability and national digital-infrastructure tailwinds that mature hubs such as Singapore and Tokyo struggle to provide, McCarthy said.
That physical runway matters for a company trying to assemble large GPU capacity rather than only sell standard cloud regions.
Pipeline Pressure Is Building
The CoreWeave project lands in a regional market already expanding quickly.
Cushman & Wakefield research put Asia-Pacific's data centre development pipeline at 26.5 GW in the first half of 2026, the highest level it has tracked.
New projects added 7.1 GW, while delivered operational capacity reached 1.4 GW.
Southeast Asia accounted for nearly half of under-construction capacity, making the region a natural test of whether AI cloud demand can support larger local campuses.
The remaining public gap is the next Asia-Pacific market after Indonesia, making the first deployment a test of how much sovereign and enterprise AI demand can translate into contracted infrastructure.
