Aolani Plans to Deploy 22,000 Nvidia Blackwell Ultra Chips, Accelerating Southeast Asia AI Data Center Expansion

TradingKey - Singapore AI cloud service provider Aolani announced a strategic computing power partnership with Nvidia (NVDA), expecting to deploy 22,000 Blackwell Ultra GPUs across two AI factories in Malaysia and the Philippines in early 2027.

Upon completing this deployment, Aolani's total GPU deployment will exceed 48,000 units, with total AI factory capacity exceeding 100 megawatts, further expanding the potential deployment scale of Blackwell Ultra in Southeast Asia.

Which Businesses Will Use 22,000 Blackwell Ultra Chips?

Aolani was founded in 2023 and joined the Nvidia Cloud Partner Program in the same year, primarily offering GPU cloud services to AI enterprises, startups, developers, and institutional clients.

Blackwell Ultra is designed for workloads such as model pre-training, post-training, inference, and agentic AI. Aolani will use these GPUs to expand its accelerated computing services in Southeast Asia, providing local clients with cloud computing resources based on the Nvidia platform.

These GPUs are expected to be deployed in early 2027. The two parties have not disclosed the procurement amount, specific system configurations, customer lists, or related contracts. Project progress will be subject to financing, equipment delivery, and data center construction.

Why Nvidia Adopts Revenue Sharing and Credit Support Models

Nvidia is adopting this model primarily to expand the deployment scale of AI infrastructure and link product sales to cloud service usage. In addition to standard product revenue, Nvidia can also share in a portion of the cloud service revenue generated by the supported computing power, increasing usage-linked recurring revenue.

Aolani stated that the two parties will adopt a revenue-sharing and credit-support model to align infrastructure deployment with customer demand. According to Nvidia's explanation of the model, AI cloud service providers purchase Nvidia infrastructure to provide computing services externally, allowing Nvidia to earn product revenue along with a portion of the cloud service revenue.

Credit support helps AI cloud service providers secure project financing, easing capital pressure during the initial stages of large-scale deployment. The two parties have not yet disclosed the financing size, credit support providers, revenue split ratios, or risk-bearing terms for this collaboration. The actual effectiveness of this model will depend on financing costs, customer demand, and GPU utilization rates.

Why Is AI Infrastructure Investment Heating Up in Southeast Asia?

It is reported that major global cloud service providers have invested a total of more than $50 billion in AI data centers and cloud infrastructure in Southeast Asia, with Malaysia being a major market where data center investment in the region is concentrated.

Google (GOOGL) plans to invest $2 billion to build its first data center and Google Cloud region in Malaysia; Microsoft (MSFT) will also invest $2.2 billion over four years to expand local cloud computing and AI infrastructure.

The data center market in the Philippines is relatively small. Aolani said its planned AI factory is expected to become one of the country's earlier large-scale AI infrastructure platforms, providing high-performance computing resources for enterprise, developer, and institutional clients.

Demand for digital services in Southeast Asia continues to grow, with multiple countries expanding power, cloud computing, and data center development. The progress of relevant projects remains subject to power supply capacity, grid connection speed, cooling conditions, data regulations, and advanced chip export policies.

Whether the Aolani project can be commissioned on schedule depends on financing, equipment delivery, and engineering construction; its commercial performance post-commissioning will depend on customer demand, service pricing, and GPU utilization rates. If the project proceeds as scheduled, it will further expand the deployment scale of Nvidia's Blackwell Ultra in Southeast Asia.