Southeast Asia Just Placed a 3.1 Billion Dollar Bet on AI Compute

By Omar aka Crypt0 · Sep 14, 2026

While the world argues about who gets to build the next frontier model, Southeast Asia is buying the shovels. Zankore, a dedicated AI compute and neocloud platform, said Wednesday it signed a senior term loan facility of up to 3.1 billion dollars to roll out Nvidia powered GPU and cloud infrastructure across Indonesia and Southeast Asia. The platform is a joint effort with Indonesian telecom operator Indosat Ooredoo Hutchison, Nokia, and Nvidia itself. Qatar's Ooredoo holds 49 percent as founding shareholder and lead investor. Citi served as exclusive debt adviser on the loan and advised Indosat on Zankore's equity raise back in August. The initial build is 100 megawatts of Nvidia AI infrastructure, scaling toward one gigawatt of Nvidia DSX AI Factory capacity, with roughly 200 megawatts of initial capacity targeted for the first half of 2027.

The loan structure matters as much as the number. This is debt, raised from bankers who looked at the demand for AI compute in Southeast Asia and decided future GPU rental revenue was solid enough to lend against. That is a milestone for the whole neocloud industry. Cloud startups used to live or die on venture rounds. Now Citi is arranging billion dollar debt facilities against GPU capacity the way banks once financed office towers. The money is moving from speculative to structural, and it is moving fastest in the regions where AI demand is growing from a smaller base.

Zankore is part of a regional pattern that took shape this month. Nvidia backed Firmus signed a multi year deal with OpenAI to supply computing capacity from two data centers in Malaysia, pushing Firmus past 900 megawatts of contracted capacity. Nvidia also announced partnerships with Australian cloud and data center firms targeting up to two gigawatts of AI capacity by 2027. OpenAI, Anthropic, and the neoclouds are building a ring of compute around the Asia Pacific, chasing power, land, and friendly governments. The map of AI infrastructure is being redrawn around wherever the electrons are cheapest and the permits arrive fastest.

For readers, the signal is that AI capacity is becoming a global commodity business, and the build out is accelerating in places the Western press rarely covers. Cheaper, more distributed compute means more AI products launching closer to more people, built for local languages, local markets, and local problems. The next breakout AI app might be trained on Indonesian GPUs before breakfast.

Sources

New to crypto? Read the crypto glossary, browse frequent questions, read our story, or explore the story archive.