- OpenAI has contracted dedicated AI compute from two Firmus sites in a multi-year Malaysia data centre deal.
- The agreement pushes Firmus past 900MW of contracted capacity and brings Nvidia’s Vera Rubin platform to Southeast Asia.
OpenAI has contracted dedicated AI compute capacity from two Firmus AI Factory sites in Malaysia, in a multi-year partnership the Australian infrastructure company announced from Sydney on September 8. The deal makes OpenAI an anchor customer of Firmus and lifts the company’s total contracted capacity across all customers above 900MW, according to the announcement.
Firmus did not disclose the locations of the two Malaysian sites, the megawatts allocated to OpenAI, or the value of the contract. Asked by Tech Wire Asia about the site locations, the Malaysian share of contracted capacity, grid supply arrangements and export licensing for the Nvidia systems, the company said it had no further comment beyond the announcement.
The announcement, however, did indicate its broader portfolio now spans seven AI factories across Australia, Singapore, Indonesia and Malaysia. Two are operational, in Australia and Singapore, and five are under development with a stated target of ready-for-service within 24 months. Firmus was valued above US$10.5 billion in its last fundraising, backed by Nvidia, Jane Street, Blackstone funds and Coatue Management, and is widely reported to be preparing an Australian listing.
What the Malaysia data centre sites will run
Firmus said it will deploy Nvidia’s Vera Rubin platform at scale across Asia-Pacific as part of the partnership. The facilities will pair Nvidia Vera Rubin NVL72 rack-scale systems with the company’s own HyperCube platform, which combines liquid cooling, mechanical systems and electrification into a repeatable unit. Firmus said the HyperCube is manufactured and prefabricated in regional New South Wales.
The design work, the manufacturing and the prefabrication sit in Australia. What Malaysia contributes to the arrangement is land, grid connection and water, which is the same division of labour that has characterised most of the country’s data centre boom since 2023. Malaysia has become the region’s fastest-growing data centre market, and the buildout has drawn steady scrutiny over electricity and water consumption, particularly in Johor.
Firmus also said it intends to establish an Australian AI Access Program to give eligible researchers and organisations access to compute for work in science, education, agriculture, energy and climate resilience. No equivalent programme was announced for Malaysia, despite the Malaysian sites being the capacity the OpenAI contract is built on.
Framing versus footprint
Firmus co-founder and co-CEO Tim Rosenfield said the partnership marks “the moment Asia-Pacific becomes a producer of intelligence” rather than a consumer of it, adding that the company had spent seven years building a footprint running from Southeast Asia to Australia.
Sachin Katti, OpenAI’s VP of compute strategy, said the Malaysian facilities would “help us serve growing demand for OpenAI’s products across the region” and globally. On the announced terms, a US frontier lab is renting capacity from an Australian operator running Australian-built hardware systems on Malaysian soil.
Whether that constitutes producing intelligence or hosting someone else depends on questions the announcement leaves open, including how much of the capacity will be available to Malaysian enterprises and researchers once OpenAI’s allocation is met.
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.
Tech Wire Asia is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
