Indian firm seeks $3 billion to buy next-gen Nvidia chips

The $3 billion debt play

AM Intelligence, a firm based in Hyderabad, is reportedly looking to secure up to $3 billion in debt to fund its hardware expansion. As reported by CNBC, the company intends to raise these funds through a combination of bank loans and bond issues. While the specific split between these two types of debt has not been finalized, the firm is moving quickly to secure the capital.

To bolster its financing, AM Intelligence expects to receive approximately $300 million in advance payments from potential offtakers. These offtakers are the entities that will eventually use the computing power provided by the infrastructure. This mix of debt and pre-payment highlights the high demand for specialized AI hardware in the current market.

The primary goal of this funding is to purchase Nvidia's next-generation Rubin graphics processing units (GPUs). GPUs are specialized processors designed to handle the heavy mathematical workloads required for artificial intelligence. By securing this debt, the company aims to position itself as one of the largest committed buyers of Rubin chips outside of North America.

Scaling the infrastructure

The company plans to invest a total of $6 billion to establish 100MW of computing power. This capacity will be distributed across data centers located in both India and Malaysia. The 100MW setup will specifically utilize 29,000 of Nvidia's Rubin chips to provide high-performance computing resources.

AM Intelligence does not intend to stop at 100MW. The firm has announced plans to add another 300MW of capacity to the market within the next 15 months. Achieving this scale will require an investment exceeding $20 billion. This rapid expansion suggests a goal of capturing a significant portion of the growing AI infrastructure market in the region.

The company is also looking at a much longer timeline for its global footprint. A spokesperson for the firm noted that they are developing 5 GW of AI data centers and 1 GW of compute-as-a-service across the United States, Europe, and India. Within two years, they aim to have 400MW of compute capacity fully operational.

A shift in AI needs

The choice of Nvidia's Rubin chips is significant because they are specifically tailored for agentic AI. Agentic AI refers to systems capable of taking actions to achieve complex goals, rather than just generating text or images. This hardware choice suggests that AM Intelligence is preparing for a specific type of enterprise demand.

According to Neil Shah of Counterpoint Research, the market is shifting toward running low-latency production agents. While previous efforts focused on training raw foundational models, many enterprises are now looking for ways to run models in real-time environments. The Rubin GPUs are expected to provide a first-mover advantage for the company in the Indian market because of this specific technical focus.

For comparison, other firms in the region are pursuing different hardware paths. For example, Mumbai-based Yotta Data Services plans to invest $2 billion in an AI hub using Nvidia's Blackwell Ultra GPUs. While Blackwell Ultra is often used for the heavy lifting of model training, Rubin is being positioned for the next phase of production use.

The rise of neoclouds

AM Intelligence is part of a broader trend where Indian energy companies are transforming into what are known as neocloud companies. These companies set up data centers with massive computing power that can be leased out to hyperscalers, which are large-scale cloud service providers like Amazon or Google. This transition allows energy firms to diversify their business models by leveraging their existing power infrastructure.

Several major Indian players are already making moves in this space. Key examples include the Adani Group, Reliance, and Tata Power. By building the physical infrastructure for AI, these companies can provide the backbone for the digital economy. This creates a vertical integration where the companies providing the electricity also provide the data centers and the chips.

AM Intelligence benefits from this trend by being owned by the founders of Greenko, a leading Indian renewable energy firm. This connection provides a strategic advantage in securing the power necessary to run high-density data centers. The move reflects a broader industrial strategy to turn India into a global hub for AI infrastructure and services.