© The Indian Express Pvt Ltd
Tags:
Journalism of Courage
Google AI Data Centre in Visakhapatnam: Google has announced an investment of $15 billion over five years to set up an artificial intelligence (AI) data centre in Andhra Pradesh, in what will be its biggest investment yet in the country. The tech giant has partnered with the Adani Group and Airtel to build the infrastructure for the project, which also includes the construction of a new international subsea gateway.
The big investment commitment comes amid soured ties between India and the United States, with Prime Minister Narendra Modi having called for people to prioritise local, or swadeshi, products, with key members of his Cabinet promoting domestically made communication and navigation apps.
Google’s data centre will come up in the port city of Visakhapatnam, and will be a part of the company’s global network of AI centres spread across 12 countries. Here’s a lowdown on what Google’s plans with the data centre are, and some policy imperatives on whether India should be incentivising these compute facilities.
But first, a distinction between traditional and AI data centres:
An AI data centre is different from a traditional data centre in terms of it being specifically tailor-made to support AI applications. Traditional data centres typically consist of CPU-powered servers that support tasks such as website hosting, cloud storage, and business applications.
An AI data centre, on the other hand, is needed to process vast amount of data, and is equipped with more cutting edge hardware such as GPUs to handle tasks like image and video production, generative AI, etc. Since these data centres handle high-performance GPUs, they require robust power infrastructure and cooling solutions, making them significantly more energy-intensive than traditional data centres.
“According to an analysis conducted by Access Partnership and commissioned by Google, the AI hub is expected to generate at least $15 billion over five years (2026-2030) in American gross domestic product (GDP) because of new economic activity from increased cloud and AI adoption, as well as the American talent and resources involved in developing and operating the AI hub. This initiative creates substantial economic and societal opportunities for both India and the United States — while pioneering a generational shift in AI capability,” Google said in a blog post.
Being developed with AdaniConneX and Airtel, the AI hub will be built with the same infrastructure that powers Google products like Search, Workspace, and YouTube. AdaniConneX is a joint venture between Adani Enterprises and EdgeConneX, set up to develop and operate data centers throughout India. AdaniConneX will provide green energy services to the AI hub, which claims to be run entirely on clean energy.
Google’s AI hub investment includes construction of a new international subsea gateway, including multiple international subsea cables to land in Visakhapatnam on India’s eastern coast — connecting to Google’s more than two million miles of existing terrestrial and subsea cables. Airtel will assist Google on this.
Google will work with local partners to deliver new transmission lines, clean energy generation, and energy storage systems in Andhra Pradesh. “The project also builds on both companies’ commitment to sustainability, and will see co-investment in new transmission lines, clean energy generation and innovative energy storage systems in Andhra Pradesh. This will not only support the data centre’s operations but also enhance the resilience and capacity of India’s electricity grid,” the Adani Group said in a statement.
India’s data centre market is currently estimated to be worth $10 billion, with around $1.2 billion revenue generated in FY24, according to a recent report by Anarock. As per real estate firm JLL, India is expected to add 795 MW of new capacity by 2027, taking the total capacity to 1.8 GW.
However, within sections of the government, there are some concerns on whether India should offer policy incentives for data centres, especially given how energy-intensive this sector is and the high capital-low employment intensiveness of this industry. Andhra Pradesh Chief Minister N Chandrababu Naidu, an NDA ally in the Centre and the state, said Google’s data centre will directly or indirectly create 1.88 lakh jobs.
AI data centres have a massive energy demand. As per the International Energy Agency (IEA), data centre power usage could double by 2026, making the challenge for companies to become net zero or carbon negative by 2030 increasingly unattainable.
Power consumption and related infrastructure costs are a key part of a data centres’ capital and operational expenditure. According to CareEdge Ratings, 40 per cent of capex of such facilities goes towards electrical systems and 65 per cent of operating cost is on account of electricity consumption. It takes between Rs 60-70 crore to set up one MW of data centre capacity in India.
While renewables have been the first choice for companies, there are hurdles that green power poses. Nuclear offers a solution, especially since the grid needs new electricity sources to support AI applications and nuclear solutions offer a clean, round-the-clock power source that can tide over the limitations of renewables — not generating power when the sun’s not shining or the wind is not blowing, and absence of adequate storage options to bridge the shortfalls.
The Indian Express had earlier reported that the Indian government was open to exploring the use of nuclear energy to power the data centre boom in India, a trend which is already playing out in countries such as the US, since such centres grow exponentially owing to the rise of AI.
