Yazılar

Google announces $15 billion AI data centre in India, its biggest investment yet

Google will invest $15 billion over the next five years to build a major artificial intelligence (AI) data centre in the southern Indian state of Andhra Pradesh, marking the company’s largest-ever investment in India. The data centre, to be located in Visakhapatnam, will have an initial capacity of 1 gigawatt, serving as Google’s largest AI hub outside the United States.

The announcement came during an event in New Delhi attended by India’s finance and technology ministers. Google Cloud CEO Thomas Kurian said the project aims to support India’s growing AI ambitions: “This long-term vision we have is to accelerate India’s own AI mission.”

The investment underscores Google’s commitment to expanding its global data infrastructure, with the company planning to spend around $85 billion this year worldwide to boost cloud and AI capabilities.

However, the timing of the announcement coincides with rising diplomatic tension between Washington and New Delhi, following U.S. tariffs on Indian goods and calls within India to boycott foreign products. Despite this, Google said the initiative “creates substantial economic and societal opportunities for both India and the United States.”

Google will partner with Adani Group and Airtel to build the new facility and its accompanying international subsea gateway, which is expected to generate over 188,000 jobs, according to earlier state estimates.

The project places Google alongside Microsoft and Amazon, who have also invested heavily in India’s rapidly expanding data centre market — home to nearly a billion internet users and a booming digital economy.

Oracle Seeks to Raise $18 Billion in Debt to Fund AI Cloud Push

Oracle is planning to raise $18 billion in debt, according to a regulatory filing on Wednesday, as it accelerates investment in cloud infrastructure to meet soaring demand from artificial intelligence clients.

The enterprise software and cloud services giant has been expanding its capital spending to deliver on major contracts, including agreements with OpenAI, which are expected to drive significant growth in its cloud business.

According to a pricing term sheet filed with the U.S. Securities and Exchange Commission, Oracle will sell the debt in six tranches.

In a separate filing, the company said proceeds could be used for general corporate purposes, including stock buybacks, debt repayment, or acquisitions, in addition to infrastructure investment.

The debt sale highlights how rising AI adoption is reshaping the priorities of major tech firms, with Oracle joining a growing list of companies tapping capital markets to finance the costly buildout of hyperscale data centers.

EcoDataCenter Secures €600 Million to Expand AI Data Centres

Swedish digital infrastructure firm EcoDataCenter announced on Tuesday that it has secured €600 million ($703.5 million) in debt financing from Deutsche Bank’s Private Credit and Infrastructure unit to accelerate its AI-focused data centre expansion.

The funding will be used to build large-scale facilities in Falun and Borlänge, north of Stockholm, aimed at handling compute-intensive AI workloads and next-generation high-performance computing.

CEO Peter Michelson, a former Ericsson executive, said the new financing provides a two-year runway, but suggested additional capital may be needed as demand for AI infrastructure accelerates.

“If we were to stop building tomorrow, we would be a highly profitable company… but we obviously have ambitions for much more than that,” Michelson told Reuters.

The deal follows €450 million raised earlier this year by EcoDataCenter’s owner Areim, bringing total financing since 2023 to €1.8 billion.

Sweden is becoming a European hub for AI data centres, with major expansions from Microsoft, Meta, and Alphabet driven by the country’s stable power grid and connectivity advantages.

EcoDataCenter, which opened its first site in 2019, counts clients such as DeepL and BMW. Last year, it partnered with CoreWeave to build one of Europe’s largest AI training clusters and now hosts a Nvidia Blackwell SuperPod for DeepL, underscoring its growing role in Europe’s AI infrastructure race.