CoreWeave says it plans to establish three AI data centres in Indonesia with a combined 360 megawatts of contracted IT power. The 4 August announcement marks the US company’s first planned data-centre presence in Asia-Pacific and targets 2028 for the facilities to come online. For Indonesia’s digital economy, the proposal could add a large new pool of specialised computing capacity, but none of the announced sites is operating today.
What CoreWeave has announced
The company describes the project as an extension of its purpose-built cloud platform for artificial-intelligence workloads. Contracted IT power is a capacity commitment for the planned facilities; it is not the same as energised infrastructure, installed computing hardware or computing service available to customers.
A 360 MW plan would be material in a market competing to host AI development, cloud services and data-intensive industries. Local capacity can reduce the distance between regional users and computing resources, while potentially widening choices for companies that need high-performance training or inference infrastructure.
Why the scale matters to Indonesia
The proposal also raises an infrastructure delivery test. Large data centres depend on grid connections, cooling, land, network links and reliable power. CoreWeave’s release does not identify the three locations, disclose an investment value or explain the planned electricity mix, so the project’s economic and environmental effects cannot yet be measured from the announcement.
What has to happen before 2028
The 2028 date is an expectation, not a completed construction schedule. Permits, site works, equipment procurement, grid availability and customer demand can all affect when contracted power becomes usable computing capacity. SEA Connect therefore treats the announcement as a forward infrastructure plan rather than a current operational expansion.
For Indonesian technology companies and regional buyers, the useful milestones will be specific: confirmed sites, construction starts, grid and network agreements, hardware deployment, customer availability and published service regions. Those steps would show whether the proposal is moving from contracted capacity into a functioning local AI-cloud market.
Source note
This analysis uses CoreWeave’s 4 August investor release as the primary source. It does not repeat an ordinary-media framing or present the company’s targets as independently verified delivery. The article is a transparent follow-up on a recent infrastructure plan, not breaking news dated to 10 August.