It is reported that an artificial intelligence infrastructure company in India is planning to raise debt financing to purchase NVIDIA Rubin graphics processors.
Headquartered in Hyderabad, AM Intelligence plans to raise funds through bond issuance and bank loans to purchase NVIDIA's new generation of chips. Reports indicate that the financing ratio between bonds and bank loans has not yet been finalized, but the company expects to receive a prepayment of $300 million from potential buyers of computing power.
AM Intelligence claims to be one of the largest "locked-in purchasing customers" of Rubin GPU outside of North America, but the company has not responded to requests for comments.
The company was founded by the founder of Greenko ( Greenko ), a leading renewable energy enterprise in India. In a press release issued on October 5th, AM Intelligence stated that it plans to invest $6 billion to build a computing power capacity of 100 megawatts, which will be deployed in multiple computing centers in Malaysia and India.
The company stated that in order to support this 100 megawatt of computing power, they will purchase a total of 29,000 NVIDIA Rubin chips; and plan to add another 300 megawatts of computing power within the next 15 months, with a total investment of over 20 billion US dollars.
More and more Indian AI infrastructure companies are starting to offer compute power as a service ( Compute-as-a-Service ), and AM Intelligence is one of them. Earlier this year, Jota Data Services ( Yotta Data Services ) in Mumbai announced that it would use NVIDIA Blackwell Ultra GPU to build a AI compute power hub.
Neil Shah, vice president of market research firm Counterpoint Research, stated that several Indian energy companies are transforming into new types of cloud providers, building data centers that can be rented to hyperscale cloud service providers; Adani Group, Reliance Industries, and Tata Power are typical examples.
Shah stated that NVIDIA Rubin GPU is specifically designed for the agent AI ( agentic AI ), which will give AM Intelligence a first-mover advantage in India.
By next year, the focus of enterprises will shift to "operating low-latency production-grade AI intelligents," rather than just the basic large model training tasks that Blackwell Ultra GPU emphasizes.
A spokesperson for the company stated in a document that AM Intelligence is building data centers with a total capacity of 5 gigawatts in India, the United States, and Europe, as well as providing 1 gigawatt of computing power as a service; they plan to achieve a computing power capacity of 400 megawatts within two years.












