Firmus Secures OpenAI Capacity Deal in Malaysia
Nvidia-backed data center startup Firmus has signed a deal to supply OpenAI with AI compute capacity in Malaysia, according to a September 8, 2026 report. The agreement marks a significant expansion of OpenAI’s infrastructure footprint in Southeast Asia, a region racing to become a global AI hub.
Firmus, which specializes in sustainable data centers powered by Nvidia GPUs, will host OpenAI workloads at its facilities in Malaysia. The financial terms of the deal were not disclosed, but the partnership underscores the surging demand for AI compute capacity outside traditional markets like the U.S. and Europe.
Why Malaysia Is Emerging as an AI Data Center Hotspot
Malaysia has aggressively courted tech investment, offering tax incentives and streamlined approvals for data center projects. The country’s strategic location, relatively low energy costs, and growing digital infrastructure make it an attractive alternative to neighboring Singapore, which has imposed moratoriums on new data centers due to land and energy constraints.
Firmus’s Malaysia operations are part of a broader push by Nvidia to diversify its AI ecosystem. Nvidia has invested in Firmus as part of its strategy to ensure that its GPUs power AI workloads globally, even as export restrictions limit sales to certain countries.
OpenAI’s Global Capacity Buildout Accelerates
OpenAI has been rapidly expanding its compute capacity to support the growing usage of its AI models, including ChatGPT and enterprise offerings. The deal with Firmus is one of several recent moves by OpenAI to secure data center capacity outside the U.S., following similar agreements in Europe and Asia.
This capacity expansion is critical for OpenAI to maintain its competitive edge against rivals like Google and Anthropic, who are also investing heavily in AI infrastructure. By leveraging Firmus’s Malaysia facility, OpenAI can reduce latency for customers in Asia and comply with local data residency requirements.
Nvidia’s Strategic Bet on Southeast Asia
Nvidia’s investment in Firmus is part of a larger trend of chipmakers partnering with data center operators to ensure demand for their hardware. In 2026, Nvidia has faced supply constraints for its latest GPUs, but demand remains insatiable from AI companies and cloud providers.
Malaysia is becoming a key node in Nvidia’s global network, alongside other Southeast Asian countries like Indonesia and Thailand, which are also seeing data center investments. The region’s proximity to major Asian markets and its improving power infrastructure make it a logical choice for AI compute.
What This Means for the AI Supply Chain
The Firmus-OpenAI deal highlights the growing importance of data center availability as a bottleneck in AI development. As AI models become more complex, the need for specialized compute grows, and companies are willing to strike long-term deals to secure capacity.
For investors, this deal signals continued robust spending on AI infrastructure, which bodes well for Nvidia’s revenue outlook. However, it also raises questions about the sustainability of such buildouts, given rising energy costs and environmental concerns.
Data center operators like Firmus are increasingly differentiating themselves by offering green energy solutions, which could become a key competitive advantage as regulators tighten emissions standards.
Watch for Firmus’s Next Capacity Announcements
In the coming months, watch for Firmus to announce additional capacity expansions in Malaysia or other Southeast Asian markets, which would confirm the sustainability of this partnership. The specific number to watch is the total megawatts of AI-ready capacity Firmus commits to OpenAI by early 2027.
If the deal expands to include OpenAI’s training workloads, it would signal a deeper strategic alliance. Conversely, any delays in data center construction or regulatory hurdles could test the relationship.
Comments are closed.