London-headquartered security intelligence specialist Gorilla Technology Group has last week announced that it has acquired 40 acres of land in Thailand to develop a 200 MW AI data center campus to meet growing Southeast Asian compute demand.
Located in Korat (Nakhon Ratchasima), the campus will comprise six data halls, five of which will have about 30MW capacity each and one will have about 50MW. Construction is expected to commence in July 2026 with the first phase targeted for completion by Q1 2027.
Access to power, dark fibre connectivity and water infrastructure have already been identified, with the campus having the operational flexibility to support current and next generation GPU systems, according to the firm’s press release.
Once fully deployed, Gorilla is targeting about US$1.5 billion of annualised revenue from the campus starting 2028.
The campus is designed to support Gorilla’s DBOT model, under which Gorilla designs, builds, operates and commercializes critical AI infrastructure for governments, enterprises and strategic partners.
The project is expected to create over 1,000 jobs across engineering, operations, infrastructure management and support services.
“Anyone can talk about AI, but very few can put together the actual physical platform required to run it,” said Jay Chandan, Chairman and CEO of Gorilla Technology. “This milestone deserves to be celebrated. We’ve secured the land and utilities needed for large-scale compute and cleared hurdles that many others have not.”
“Our funding strategy is centered on project level debt, infrastructure debt, asset backed financing, potential bond structures and non-dilutive capital sourced through Gorilla Technology Capital. We are targeting long duration institutional capital, including pension funds, endowments, superannuation funds and infrastructure investors that understand the real asset nature of AI compute. We are here to build a serious business using smart capital.”
Dr. Rajesh Natarajan, Group CTO of Gorilla Technology said: “The current planning model is based on 200MW facility load supporting about 150MW of net IT load and assuming a target PUE of about 1.3. Under an illustrative all-GB-300 configuration, using approximately 142kW per rack and 72 GPUs per rack, this would support approximately 76,000 GPUs at full deployment. The more important point is that the campus is being structured for phased capacity delivery, allowing Gorilla to match infrastructure deployment with customer demand and engineering readiness.”