Mint Primer | Stargate Wars: Where is India in the AI infra race?

What’s Project Stargate all about?

It’s a $500-billion joint venture with OpenAI, Oracle, SoftBank, and MGX to build AI-focused infrastructure including data centres across the US over the next four years—Trump’s first major business initiative since returning to office. The project is aimed at countering China’s AI advance, stressing the role of geopolitics in tech policies and strategies. The new company will initially deploy $100 billion. SoftBank will fund the venture, while OpenAI will oversee operations. Key partners include Arm, Microsoft, Nvidia, and Oracle. Construction is underway in Texas, with additional sites under review.

Read more: Honey, I’m home: Is this the era of domestic robots?

Who leads in the US vs China AI race?

The US leads the global AI ecosystem, excelling in research, machine learning models, private investment, and job creation. However, while China-based Baidu’s Ernie bot may not match OpenAI’s ChatGPT, China ranks second in AI, as per Stanford University rankings. China attracted $67.2 billion in AI-related investment in 2023 compared to $7.8 billion for the US, and is way ahead in AI patents and notable machine learning models (61 vs 15). China is also innovating fast with AI lab DeepSeek’s R1 open-source reasoning model rivalling OpenAI’s o1 in performance across reasoning tasks and costing less.

Any other thinking behind Stargate?

Nvidia, the world’s largest AI firm, relies on Taiwan Semiconductor Manufacturing Company for chips. China views Taiwan as a breakaway province. Governments are prioritizing AI-ready data centres, with generative AI projected to account for 40% of advanced-AI workloads. Cloud providers drive this demand, underscoring the critical role of infrastructure in AI.

What are others doing about AI infra?

The UK ranks third in AI, with plans for significant data centre expansion to attract AI firms, followed by India and the UAE. India is strengthening its AI infrastructure, including RackBank’s 80MW AI-focused data centre in Madhya Pradesh. India has around 150 data centres, including AWS, Microsoft Azure, and Google Cloud, and its IndiaAI Mission, with a ₹10,372 crore budget, focuses on enhancing computing infrastructure, supporting deep-tech startups, and fostering AI development.

What more does India need to do?

India will build AI infrastructure using 10,000 GPUs (graphics processing unit) through public-private partnerships but it must address restrictions on US AI chip exports. Ajai Chowdhry, founder of HCL and chairman of the EPIC Foundation and National Quantum Mission of India, says Stargate shows the “serious intent” of the US to take control of AI and own all large language models (LLMs). Dubbing Stargate as the “weaponisation of tech”, he believes India must create its “own AI doctrine” by controlling its data.

Read more: Companies expected to increase Gen AI investments by half this year