OpenAI is expanding its global artificial intelligence infrastructure footprint into Malaysia through a major multi-year agreement with Firmus, an Australia-based AI infrastructure company backed by NVIDIA.
The agreement will give OpenAI computing capacity across two Firmus data centers in Malaysia, making the ChatGPT maker an anchor customer as demand for the infrastructure required to train and run advanced AI models continues to surge.
The deal also pushes Firmus’ contracted capacity across its customer base above 900 megawatts, highlighting the extraordinary scale of the global AI data-center buildout.
OpenAI Becomes an Anchor Customer for Firmus
Firmus announced on September 8 that OpenAI had signed a multi-year agreement for computing capacity from two Malaysian data centers.
The arrangement is strategically important for Firmus because securing OpenAI as an anchor customer provides a major long-term demand source for its expanding AI infrastructure portfolio.
For OpenAI, the agreement provides additional computing capacity without requiring the company to build every facility itself.
That distinction is increasingly important as AI companies race to secure electricity, land, cooling systems, networking infrastructure and advanced accelerators.
The AI infrastructure business is rapidly evolving from individual server purchases into massive long-term capacity commitments.
Firmus’ Contracted Capacity Surpasses 900 MW
The OpenAI agreement takes Firmus’ contracted capacity across its customer portfolio to more than 900 MW.
That is an enormous amount of planned AI computing infrastructure.
Firmus describes its facilities as AI factories, designed around the high-density computing requirements of modern artificial intelligence workloads.
Its portfolio includes NVIDIA-powered infrastructure, making the company part of the broader ecosystem supporting the expansion of GPU-based AI computing.
The 900 MW figure also illustrates how quickly AI infrastructure companies are scaling.
A few years ago, data-center expansion was largely discussed in terms of individual facilities.
Today, AI infrastructure deals increasingly involve hundreds of megawatts of contracted capacity.
Why Malaysia Is Becoming an AI Data-Center Hub
Malaysia has emerged as one of Southeast Asia’s fastest-growing data-center markets.
Its geographic position gives operators access to major Asian markets, while the country has attracted significant investment from global technology companies.
But the rapid growth is also putting pressure on the country’s electricity infrastructure.
Reuters reported that Malaysian data centers consumed a record 9.3% of total electricity use in mid-August 2026, compared with an average of around 7% for the year. The share could potentially rise as the industry expands.
That makes OpenAI’s latest deal significant beyond the technology industry.
It adds another major AI workload to a country already dealing with rapidly increasing data-center power demand.
OpenAI Needs More Compute Around the World
OpenAI’s demand for computing capacity continues to rise as its AI products become more capable and more widely used.
The company is no longer building infrastructure only for traditional chatbot interactions.
Modern AI workloads include:
- Large-scale model training
- AI inference
- Coding agents
- Autonomous AI systems
- Image and video generation
- Enterprise AI
- Deep research
- Real-time AI applications
Each of these workloads requires enormous amounts of computing power.
As models become larger and AI agents perform longer and more complex tasks, infrastructure requirements can increase further.
That is why OpenAI has been building relationships with multiple infrastructure providers instead of depending on a single data-center operator.
NVIDIA’s Role Makes the Deal Even More Interesting
Firmus is backed by NVIDIA, making the OpenAI agreement particularly interesting in the context of the AI hardware ecosystem.
NVIDIA remains one of the most important suppliers of AI accelerators and complete computing systems used in large-scale data centers.
Firmus’ facilities use NVIDIA technology, meaning OpenAI’s new Malaysian capacity is part of the broader NVIDIA-powered AI infrastructure ecosystem.
This does not mean NVIDIA is directly operating OpenAI’s Malaysian data centers.
Instead, it demonstrates how NVIDIA’s hardware ecosystem extends through specialized infrastructure providers that build and operate large AI computing facilities.
Firmus Is Building a Global AI Infrastructure Network
The Malaysia agreement is not an isolated expansion for Firmus.
The company already has projects and partnerships across the Asia-Pacific region.
Its broader infrastructure footprint spans markets including Australia, Singapore, Indonesia and Malaysia, while the company has been developing large-scale NVIDIA-powered AI facilities in the region.
Firmus has also been pursuing major projects in Indonesia.
One planned project in Batam is designed around a 360 MW AI data-center campus, with infrastructure capable of supporting a very large number of NVIDIA GPUs.
That gives Firmus a potential regional advantage.
Instead of operating a single data center, the company is building a network of AI infrastructure across multiple locations.
OpenAI Is Becoming a Major Customer for AI Infrastructure Companies
The Firmus agreement also reflects a larger trend in the AI industry.
OpenAI is increasingly becoming an anchor customer for companies whose entire business is built around supplying AI compute.
That relationship can benefit both sides.
For infrastructure providers, OpenAI offers predictable demand from one of the world’s largest AI companies.
For OpenAI, long-term contracts can help secure computing capacity while avoiding the need to independently develop every aspect of data-center infrastructure.
The model resembles the way cloud providers historically worked with large enterprise customers — except the scale of today’s AI workloads is dramatically larger.
Southeast Asia Is Becoming More Important to the AI Race
The Firmus agreement also highlights Southeast Asia’s growing importance in global AI infrastructure.
The region is attracting data-center investment from major technology companies because of its growing digital economy, strategic location and access to emerging markets.
Malaysia in particular has become a major destination for data-center development.
But the region faces the same challenges as other AI infrastructure hubs.
Large AI data centers require enormous amounts of electricity and sophisticated cooling systems.
As demand grows, governments must balance technology investment against electricity availability, water consumption, land use and environmental considerations.
The Power Challenge Could Become Just as Important as the Chips
AI infrastructure discussions often focus on GPUs and AI accelerators.
But electricity is becoming an equally important constraint.
A data center can have access to the latest NVIDIA hardware and still struggle to operate if sufficient power is unavailable.
Malaysia’s rapidly increasing data-center electricity consumption demonstrates this challenge.
Reuters reported that under continued growth, data centers could potentially account for as much as 31% of Peninsular Malaysia’s electricity consumption by 2035.
That makes energy planning an increasingly important part of the AI infrastructure race.
The competition is no longer simply about who can obtain the most GPUs.
It is also about who can secure the electricity and physical infrastructure needed to run them.
OpenAI’s Malaysia Expansion Fits Its Larger Infrastructure Strategy
OpenAI has been pursuing a broad strategy of securing computing capacity through infrastructure partnerships.
The company is working with multiple providers and locations rather than relying exclusively on one geographic region.
That diversification can provide several advantages.
It can reduce dependence on individual facilities, improve geographic redundancy and allow OpenAI to access additional power and computing resources as demand grows.
Malaysia adds another strategic location to that network.
For Southeast Asia, meanwhile, having OpenAI as a major AI infrastructure customer could further establish the region as an important part of the global AI economy.
What the Firmus Deal Means for the AI Industry
The OpenAI-Firmus agreement is about much more than two Malaysian data centers.
It illustrates how the AI race is increasingly becoming an infrastructure race.
The companies building the next generation of AI systems need:
Models + Chips + Data Centers + Power + Networking + Cooling
Every part of that chain must scale simultaneously.
Firmus is attempting to become one of the companies providing the physical infrastructure.
OpenAI needs that infrastructure to expand its AI products.
NVIDIA supplies a major portion of the computing technology powering those facilities.
And Malaysia is becoming one of the locations where this entire ecosystem is being built.
OpenAI’s Malaysian Expansion Shows How Big the AI Infrastructure Boom Has Become
OpenAI’s multi-year agreement with NVIDIA-backed Firmus gives the company access to computing capacity across two Malaysian data centers and makes OpenAI an anchor customer for the infrastructure provider.
The agreement also pushes Firmus’ contracted capacity above 900 MW, showing the extraordinary scale of today’s AI infrastructure requirements.
For Malaysia, the deal strengthens its position as a major Southeast Asian AI data-center hub — while also increasing pressure on its electricity infrastructure.
For Firmus, OpenAI provides a powerful customer as the company expands across the Asia-Pacific region.
And for OpenAI, the agreement is another piece of a much larger strategy: secure enough computing capacity to support an AI ecosystem that is becoming more demanding every year.
The next phase of the AI race may therefore be decided not only by who builds the smartest models, but by who can secure enough power, chips and data-center capacity to run them at global scale.
FAQ
What did OpenAI and Firmus announce?
Firmus announced a multi-year agreement with OpenAI to provide computing capacity from two data centers in Malaysia. OpenAI will serve as an anchor customer for the facilities.
Who is Firmus?
Firmus is an Australia-based AI infrastructure company that develops and operates large-scale computing facilities for AI workloads. The company is backed by NVIDIA and other investors.
How much computing capacity does Firmus have?
The OpenAI agreement pushes Firmus’ contracted capacity across customers to more than 900 megawatts.
Why is OpenAI expanding into Malaysia?
Malaysia is becoming an important data-center hub in Southeast Asia, providing additional infrastructure capacity for companies that need large amounts of computing power.
Are Firmus’ data centers powered by NVIDIA technology?
Firmus’ AI infrastructure includes NVIDIA-powered systems, making NVIDIA an important part of the company’s computing infrastructure ecosystem.
Is OpenAI building its own Malaysian data center?
The announced arrangement is for OpenAI to use computing capacity from Firmus’ Malaysian data centers. It does not mean OpenAI is independently building and operating the facilities itself.
Why is Malaysia important for AI infrastructure?
Malaysia has become one of Southeast Asia’s fastest-growing data-center markets. Its strategic location and growing digital economy have attracted major technology infrastructure investment.
Is Malaysia facing an electricity challenge from data centers?
Yes. Reuters reported that data centers accounted for a record 9.3% of Malaysia’s electricity use in mid-August 2026, compared with an average of about 7% for the year.
What does the deal mean for NVIDIA?
The agreement demonstrates how NVIDIA’s AI computing ecosystem is expanding through infrastructure operators such as Firmus. It does not mean NVIDIA is directly operating OpenAI’s Malaysian facilities.
What does this deal mean for OpenAI?
It gives OpenAI additional long-term computing capacity as the company continues expanding its AI models, products and agentic workloads around the world.
Could Malaysia become a major AI hub?
Malaysia is already emerging as a major Southeast Asian data-center destination. Continued investment from AI companies and infrastructure providers could strengthen its position, although electricity, cooling and sustainability requirements will remain important constraints.
