Sify Technologies to Invest $5 Billion in AI-Driven Data Centers Across India

Sify Betting Big On India, To Invest $5 Bn In Data Centres: Report

The expansion aligns with India's AI mission, which has allocated INR 10,372 Cr for building AI infrastructure Cloud infrastructure provider Sify Technologies is reportedly aiming to invest $5 Bn in India over a period of five years. The company will be using these funds to build smaller AI inferencing facilities in 20 tier-II cities. Sify offers a comprehensive suite of services including telecom services, data centre services, cloud and managed services, and digital transformation solutions. It operates across more than 1,600 cities in India and runs 11 data centres. As per ET's report, citing Sify chairman and managing director Raju Vegesna, India is expected to be among the largest markets of AI services and inferencing rather than training. "We are going to invest in data centres not only in the 6-7 metros but in the next 10-20 tier-2/tier-3 cities like Lucknow, Chandigarh, Nagpur with small capacities of around 10 megawatts to create these AI factories," Vegesna said. This initiative comes on the premise that Special Economic Zones are moving to tier II cities. All IT resources and IT outsourcing companies are also expanding there. "We need to take the current IT resources to the next level with AI," he added. Sify Technologies is also looking to expand data centres, modernise internal systems with AI Ops, its cloud platforms and GPUs. With this it aims to compete with large data centre and telecom incumbents. Last year, they announced GPU-as-a-service last year and are currently running several proof-of-concepts. In September 2023, it raised INR 600 Cr from the Kotak Data Centre Fund. The company's expansion aligns with India's AI mission, launched in July 2023 with an INR 10,372 Cr allocation. The government has earmarked INR 4,563 Cr for building compute capacity of 10,000 GPUs under this mission. Another INR 1,971 Cr has been set aside for the new IndiaAI Innovation Centre.

Sify earmarks $5 billion to expand data hubs, add AI ops and GPUs

Sify Technologies plans to build smaller AI inferencing facilities in 20 tier-II cities as India is expected to be among the largest markets of AI services and inferencing rather than training, Vegesna said who is helming the only Indian company listed on Nasdaq besides Infosys.Seizing on the AI opportunity, cloud infrastructure and services leader Sify Technologies has earmarked $5 billion investment over the next five years to expand data centres, modernise internal systems with AI Ops, acquire GPUs and compete with large data centre and telecom incumbents, Raju Vegesna, its chairman and MD, said in an interview to ET. The company plans to build smaller AI inferencing facilities in 20 tier-II cities as India is expected to be among the largest markets of AI services and inferencing rather than training, Vegesna said who is helming the only Indian company listed on Nasdaq besides Infosys. What are Sify's plans going forward? Sify started as the first private internet service provider. Then, we deployed enterprise network and cloud, followed by digital transformation. In 2025, we start the journey of AI innovation at the core. Our concept is - all the three Sify entities across network infrastructure, data centres and colocation, and digital services, are going to be driven by AI. We have a potential roadmap to invest $5 billion in the next 5 years in our major data centre, tier-2 data centres, our network expansion, and also our cloud platform and GPUs. We are going to invest in data centres not only in the 6-7 metros but in the next 10-20 tier-2/tier-3 cities like Lucknow, Chandigarh, Nagpur with small capacities of around 10 megawatts to create these AI factories. How do you plan to fund the expansion? We are going to take the money when we need it. Recently we raised Rs 600 crore from the Kotak Data Centre Fund, which was backed by investors like ADIA, GIC. For me, raising $5 billion in equity plus debt is not a challenge. What is the reasoning behind expanding to 20 tier-2/tier-3 cities? Special Economic Zones are moving to the tier-2 cities. All the IT resources, IT outsourcing companies, are also going into the tier-2 cities.We need to take the current IT resources to the next level with AI. To build that, we need to invest in those cities. We want to build this digital infrastructure suitable for AI into those cities. We are operating in four verticals -- BFSI, manufacturing, retail and health sciences. Some of these require low latency, so we have to create localisation for faster reaction.AI training in a metro city is where you require large investments. But, inferencing is cost-effective which you can build in tier-2 cities. How do you see AI adoption in India? India is still taking time to adopt AI, but the opportunity is big. We need to have good data first. People started investing in the last 3-4 years, collecting good data, and the government support with the AI Mission are excellent plans. We announced the GPU-as-a-service last year, now we have some proof-of-concepts running. The way that IT helped in creating the technology and the resource pool--we are now the biggest IT hub of the world--I hope India will become the AI hub of the world. I live in Silicon Valley. What are the two buzzwords there? AI and India. People are asking, what is your AI plan? What is India plan? What are the top challenges you see at the intersection of AI and data centres? Data centres are not a real estate game, it's a technology game. And with AI being deployed, it's going to become more complex. AI requires more power -- how do you do liquid cooling? How do you get renewable energy? To run these data centres 24/7, 365 days, there are thousands of variables that could go wrong. It's not humanly possible to monitor that, so how do you use AIOps to predict failures? How do you ensure seamless connectivity to ensure it is accessible everywhere? What is your opinion on the debate about a 70-hour work week? I have mixed opinion on that. Yes, Indians need to work hard. No doubt. Ten years from now, we are going to compete with the US, China and Japan. These countries did not develop overnight. The US suffered, after the First World War, and rebuilt all infrastructure. And then Japan, after the Second World War, how they build it, how China is being built. Certainly, yes, if you want to position India into that kind of category, we have to work very hard. But the world has also changed. Instead of working hard time-wise, we have to work smarter.