Aolani, a Singapore-founded neocloud powering AI growth, has signed Letters of Intent to take up 500 megawatt (MW) of AI data centre capacity in Malaysia, with the capacity targeted to come online from the second half of 2027 across multiple sites. The announcement, issued through Media OutReach Newswire, matters less for the raw megawatts than for what it reveals about how AI infrastructure competition is shifting: from securing land and power to wiring compute into national and enterprise ecosystems.
The sites, designed to Tier III standards, will house purpose-built AI data centres supporting Aolani's expanding portfolio of AI factories and GPU deployment. Under the arrangement, Aolani will provide the AI infrastructure, GPU deployment and operating requirements across the facilities upon completion. With this commitment, the company projects it will commit up to 580MW of capacity in Malaysia by 2027, up from approximately 80MW already committed in 2026 — a more than sevenfold increase that would make Malaysia the anchor of its regional footprint.
The scale of the intended expansion carries implications beyond Aolani's own balance sheet. Malaysia has emerged as a key destination for AI infrastructure investment in the region, driven by competitive power costs, available land, and growing governmental and business support for digital infrastructure development. Those advantages have drawn hyperscalers and data centre operators, but the next competitive layer is whether that capacity serves local users rather than only exporting compute. Aolani says the expanded capacity will serve demand from global hyperscalers and AI-native companies while also making AI compute accessible locally, and it is exploring initiatives and partnerships to provide high-performance compute capacity to Malaysian sovereign AI programmes, local enterprises, researchers and start-ups.
That local-access dimension is where the announcement gains strategic weight. Countries building sovereign AI programmes need domestic or near-domestic compute to train and run models under their own data governance rules; enterprises and researchers need affordable access without navigating foreign cloud regions. A neocloud positioning itself as "hyperlocal" — Aolani's own description — is effectively betting that proximity and local partnership will matter as much as raw GPU supply. Nicholas Chia, Chief Executive Officer at Aolani, framed Malaysia as a vital strategic anchor and a proven home market, saying the country is already firmly established as a global hub for AI infrastructure, but that the next stage is using local compute to catalyse development of the local AI ecosystem and the broader economy.
The timing also matters. Capacity arriving from the second half of 2027 places this expansion in the window when many AI infrastructure projects announced in the region are expected to be energised, raising questions about power availability, grid readiness and execution. The Letters of Intent are non-binding and remain subject to negotiation and execution of definitive agreements, so the 500MW figure represents intent rather than secured capacity. Still, the direction is clear: a company that committed roughly 80MW in 2026 is signalling a 2027 pipeline seven times larger, concentrated in a single market it calls home. More details on the company's platform are available at https://www.aolanicloud.com/.
