On September 8, 2026, Firmus said OpenAI agreed to contract dedicated AI computing capacity from two Malaysian AI Factory sites. Firmus plans to deploy Nvidia’s Vera Rubin platform at scale across Asia Pacific, underscoring how access to chips, power, cooling and data centre space has become a strategic constraint f...
Published byEdited with GPT-5.6 TerraImages generated with GPT Image 2
Research answer

Create a landscape editorial hero image for this Studio Global article: What did Australia’s Nvidia-backed AI infrastructure firm Firmus announce on September 8, 2026, about its multi-year agreement to supply Ope. Article summary: Firmus announced that OpenAI will contract dedicated AI-computing capacity under a multi-year strategic partnership from two Firmus AI Factory sites in Malaysia, making OpenAI an anchor customer. The companies did not di. Topic tags: general, news, general web. Style: premium digital editorial illustration, source-backed research mood, clean composition, high detail, modern web publication hero. Use reference image context only for broad subject, composition, and topical grounding; do not copy the exact image. Avoid: logos, brand marks, copyrighted characters, real person likenesses, fake screenshots, UI text, readable text, watermarks, charts with fake numbers
OpenAI is expanding its compute supply into Malaysia through a multi-year partnership with Australian AI-infrastructure company Firmus. The agreement covers dedicated capacity at two Firmus AI Factory sites and makes OpenAI an anchor customer for the company’s Malaysian expansion. 17
Firmus and OpenAI announced the partnership on September 8, 2026. Under the arrangement, OpenAI will contract dedicated AI compute from two Firmus sites in Malaysia. Firmus said the addition of OpenAI takes its aggregate contracted customer capacity above 900 megawatts (MW). 17
An anchor customer is significant because it provides a major early commitment for new infrastructure. Firmus said OpenAI’s commitment will support deployment of additional capacity across its AI Factory platform. 17
Neither company disclosed the contract value, the amount of capacity OpenAI is buying, pricing, or a delivery schedule. Those omissions make it difficult to assess the agreement’s direct revenue impact or compare its scale with other OpenAI infrastructure commitments. 3
17
Firmus said it intends to deploy Nvidia’s next-generation Vera Rubin accelerated-computing platform at scale across Asia-Pacific. Reporting on the announcement described the Malaysian facilities as using Vera Rubin NVL72 rack-scale systems alongside Firmus’ HyperCube infrastructure, including liquid cooling and electrification. 17
3
That combination matters because advanced AI infrastructure is not simply a matter of acquiring processors. It also requires power delivery, heat management, networking and physical data-centre capacity that can support dense compute systems.
The deal gives Firmus a high-profile customer as the company expands regionally. Reuters reporting said the agreement came ahead of a rumoured Australian initial public offering and described Firmus as Nvidia-backed, with a reported valuation above $10.5 billion and investors including Nvidia, Jane Street, Blackstone-managed funds and Coatue Management. The reporting did not provide a valuation methodology or details of each investor’s holding. 3
For Firmus, the more concrete milestone is the more-than-900-MW figure: it represents contracted capacity across customers, rather than necessarily capacity that is already operating. Firmus has said its network spans Australia, Singapore, Indonesia and Malaysia, with two AI factories operating and five in development. 17
3
The partnership adds to the competition among AI companies to secure the physical inputs behind increasingly capable models: chips, electricity, cooling systems and data-centre space. Reuters framed the agreement against the wider contest among U.S. and Chinese AI groups for infrastructure capacity. 3
For OpenAI, Malaysian capacity broadens the geographic base supporting demand for its products in the region and globally, according to Firmus’ announcement. 17
Malaysia has become an important destination for data-centre investment, but rapid expansion also raises questions about electricity demand, water use and environmental effects. The available reporting does not quantify those impacts for Firmus’ two sites, so the specific local footprint of this project remains unclear. 3
The announcement is a capacity deal, not a fully priced public contract: OpenAI will be the anchor customer for two planned Malaysian Firmus AI Factory sites, and Firmus says the commitment lifts its total contracted capacity past 900 MW. Its broader importance is strategic. It links a major AI model developer, Nvidia’s next-generation platform and Southeast Asian data-centre buildout at a time when reliable compute infrastructure is becoming as consequential as the models it runs. 17
3
Studio Global AI
This page includes a source-backed answer you can continue inside Studio Global.
On September 8, 2026, Firmus said OpenAI agreed to contract dedicated AI computing capacity from two Malaysian AI Factory sites.
On September 8, 2026, Firmus said OpenAI agreed to contract dedicated AI computing capacity from two Malaysian AI Factory sites. Firmus plans to deploy Nvidia’s Vera Rubin platform at scale across Asia Pacific, underscoring how access to chips, power, cooling and data centre space has become a strategic constraint for frontier AI.