Riding high on the AI wave

India pivots from edge to hyperscale data centres, driven by a surge in AI demands and government initiatives, heralding a new data era.

For the longest time, the expansion plan for data centres in India was charted quite simply. Most industry experts pegged ‘edge’ data centres as the way that the industry would grow. Large players in the market would set up large-scale, heavy-investment hyperscale facilities that would work as the main hubs in which India’s data centres would concentrate most of their capacities.

In data centre vocabulary, ‘edge’ facilities make for data centre infrastructure that is small and regional. Inherently, data centres are large facilities that host thousands of racks and consume power in megawatts of server load. Edge data centres, on this note, are much smaller—hosting fewer racks and cabinets and consuming power in kilowatts. As a result, they came to be known as ‘hubs’. In turn, the established model for the data centre industry became a hub and spoke model.

Significantly, the expansion of India’s data centre industry is no longer reliant on edge facilities. Instead, it’s pivoting towards scaling up hyperscale units or integrating managed cloud platforms and services within existing data centre operations. This strategic shift, driven by industry dynamics, is set to redefine the trajectory of India’s data centre market.

Why the Paradigm Shift?

In a groundbreaking move in November 2022, US-based artificial intelligence (AI) research firm OpenAI, with Elon Musk as one of its initial founders and investors, unveiled ChatGPT. This revolutionary service quickly amassed 100 million users within less than 60 days since its public launch, sparking a global race to develop increasingly intelligent AI applications and creating an entire industry of generative AI.

AI, thus, came to the fore in every industry—it became commonplace in automating customer relationship management (CRM) services in enterprises, defined cyber security research operations, overlooked smart factories and global manufacturing, offered insights on weather and worldwide oil pipelines, and more.

“AI could push companies globally to spend over USD 200 billion in hyperscale data centres within two years, and the same is evident in India too.”- Jatinder Singh Pabla, Chief Sales & Marketing Officer – India, ST Telemedia Global Data Centres

As this global AI race picked up pace, the world realized one thing—this entire generation now needed more data than ever before. Today, every single task uses AI, a small subset of which is accounted for by generative AI.

And, wherever there’s AI, there’s data. The world’s servers were suddenly needed to host and crunch massive amounts of data at all times. India was no different—the erstwhile Centre issued multiple statements on prioritising AI, framed an IndiaAI Mission with an investment outlay of $1.24 billion, and hosted the Global Partnership on AI (GPAI) Summit in New Delhi to show that as one of the world’s biggest economies and an emerging market, India was ready for AI.

This, in turn, has enthused data centre operators in India and globally. More and more ventures are now entering the AI domain with cloud services and increased capacities to capture a rising share of a growing pie of data centre demand. Case in point: homegrown Hiranandani Group-backed data centre firm Yotta Infrastructure, in December last year, signed a partnership deal with global chip behemoth Nvidia for access to its in-demand H100 chips. On top of this, it also became the first domestic venture to get access to Nvidia’s ‘Blackwell’ GPUs—unveiled in March this year for AI compute purposes.

“A surge in data consumption, propelled by the proliferation of IoT devices and online content streaming, is driving the expansion of data centre capacities.”- Sridhar Pinnapureddy, Founder & CEO, CtrlS and Cloud4C

The writing was thus on the wall—AI, it seems, will transform the future of data centres too, after all.

Where Does the Market Stand Today?

On 30 April, financial services firm Goldman Sachs published a global electricity equity research mote, underlining that AI is pushing data centres to become a key contributor to rising electricity demand worldwide. While noting that AI could lead to a rise in sustainability factors too, Brian Singer, global head of investment research at Goldman Sachs, said, “The combination of AI, ex-AI increases in data demand, and a material slowdown in power efficiency gains is making data centres a critical driver of accelerating global electricity demand growth. We assume data centre power demand, excluding cryptocurrency, will grow by 160% in 2030 versus 2023—representing an increase of about 650TWh by 2030.”

Goldman Sachs’s prediction until 2030 suggests a compounded annual growth rate of 0.3% to increases in global power demand from data centres driven by AI. Because of this, data centres’ power draw could rise from 1-2% of global power consumption to 3-4%—a very significant hike. Singer added that AI will represent 19% of all data centre power demand by 2028.

In India, market stakeholders agree. “AI, the next big growth driver and enabler globally, holds enormous potential in India,” said Jatinder Singh Pabla, Chief Sales and Marketing Officer – India, ST Telemedia Global Data Centres (GDC).

“63% of Indian firms are slated to initiate investment in AI and machine learning in 2024, while 66% of all Indian CEOs see generative AI as a top investment area. India has the second-highest AI talent base in the world. With key existing and upcoming government acceleration efforts, India has an opportunity to leapfrog into global AI leadership. Globally, the resurgence in hyperscale spending in cloud and data centres is evident,” Pabla said.

“AI adoption is driving demand for data centres by increasing the need for storage, computational power, and real-time processing capabilities.”- CB Velayuthan, CEO, Digital Connexion

He added that AI could push companies to spend over USD 200 billion in hyperscale data centres in the next two years and the same is evident in India too.

Sridhar Pinnapureddy, Founder and CEO, CtrlS Data Centres, agrees that AI will “significantly impact the way data centres are designed and built.” He further explained: “The increasing adoption of cloud computing services by enterprises is a major catalyst, as data centres provide the critical infrastructure backbone for cloud service providers. Another driving force is the exponential surge in data consumption, propelled by the proliferation of smartphones, IoT devices, and online content streaming, necessitating the expansion of data centre capacities.”

CB Velayuthan, CEO of Reliance Jio and Brookfield-backed data centre firm Digital Connexion, further added, “As enterprises and consumers accelerate AI adoption, the latter is driving demand for data centres by increasing the need for storage, computational power, and real-time processing capabilities.”

The impact of AI’s rise is further reflected in the way new capacity is coming up among India’s biggest data centre operators. For instance, Equinix, which operates two active and two planned data centres in Mumbai, is expanding the latter based on demand for hyperscale facilities in India’s most populous data centre market. Multi-industry conglomerate Adani Enterprises’ data centre joint venture, AdaniConneX, and Digital Connexion, have been formed in the past two years to cash in on this demand, too.

While the market is evolving with the entry of every new player and changing technology dynamics, specific demand for edge facilities has remained slim so far. Edge facilities have yet to draw the attention that was projected, leaving AI to become the primary contributor to India’s data centre growth trajectory. 

By Vernika Awal

feedbackvnd@cybermedia.co.in