Recently, I’ve explored the exciting developments in India’s Digital Public Infrastructure (DPI), diving into innovations like Account Aggregators and the Open Credit Enablement Network (OCEN). These initiatives have showcased India’s commitment to digital transformation and financial inclusion.
As I reflected on these advancements, a question kept nagging me – What’s powering all of this digital innovation?
The answer, I realised, lies in the often-overlooked backbone of our digital world – data centres. With the rapid rise of Artificial Intelligence (AI), these data centres are becoming more crucial than ever.
So today, I want to pull back the curtain on these digital powerhouses and explore how they’re propelling India into the age of AI.
Experts predict the global AI market could hit USD 320-380 billion by 2027. And India? We’re not sitting on the sidelines. Our AI market is set to more than double, jumping from USD 7-9 billion in 2023 to USD 17-22 billion by 2027.
![](https://www.dezerv.in/blog/wp-content/uploads/2024/08/image-5.jpeg)
So, what’s driving this massive growth? Two main factors – businesses moving to the cloud and AI getting smarter by the day. But here’s the thing – none of this happens without some serious behind-the-scenes muscle. That’s where data centres come in.
Understanding Data Centres – The backbone of AI
Data centres are the engine rooms powering everything from your Netflix binge to complex AI algorithms. But what exactly goes on inside these high-tech facilities?
Imagine a warehouse filled with rows upon rows of powerful computers, all humming away 24×7. That’s essentially what a data centre is. These machines constantly crunch numbers, store information, and keep our digital lives running smoothly.
We’ve come a long way from the days when a “data centre” meant a small room with a few servers tucked away in an office corner. Today, we’re talking about massive facilities that can span several football fields. These mega data centres are like small cities, with their own power plants and cooling systems. It’s a testament to how much our digital needs have grown – and continue to grow.
But here’s the catch – all this computing power comes at a cost. Data centres are energy-hungry beasts. They gobble up about 1% of the world’s electricity. That might not sound like much, but it’s roughly equivalent to the energy consumption of a small country. And as our appetite for data grows, so does their power consumption.
Fun fact – A single ChatGPT query requires 2.9 watt-hours of electricity, compared with 0.3 watt-hours for a Google search. This means that a ChatGPT query consumes 10x the power required for a simple Google search!
What’s mind-bending here is that 10,000 Graphic Processing Units (GPUs) have to constantly work behind the scenes to train ChatGPT as they process a surreal 570 GB of text data to run this AI.
Let’s turn our attention to India for a moment. The growth we’re witnessing in this space is truly remarkable, outpacing many other markets worldwide. This trend deserves our close attention, given its potential implications for the global digital economy and investment landscape.
When it comes to data centres, India is dreaming big. We’re aiming to hit a data centre capacity of 2000 MW by 2026.
To put that in perspective, a small town might use around 50 MW of electricity. So this data centre can use as much electricity as 40 small towns! That’s a lot of power to run all the computers and keep them working properly.
The key takeaway here is that this 2000 MW data centre has an exceptional amount of computing capacity. It can effectively process and manage an immense volume of information and data. This facility is truly a powerhouse in the data centre space.
That will require an investment of about INR 50,000 crore (which translates to about USD 6.7 billion at the time of writing). It’s a lofty goal, but it shows just how serious India is about leading in the digital economy.
What is driving this substantial growth in India?
Let’s examine what’s driving this data centre boom. According to CRISIL, it’s not just one factor but a perfect storm of factors coming together.
- Growth in data consumption – The sheer size of our population means a lot of data is being created, shared, and stored. To give you an idea of the scale, mobile data traffic has been growing at a mind-boggling 20.8% year over year for the past five years.
And, just when you think it can’t get any more intense, 5G enters the picture. As this technology continues to be rolled out, it’s going to open up new possibilities we haven’t even imagined yet. But one thing’s for sure—it’s going to mean even more data needs a home.
- Tech giants staking their claim – Big tech giants (the hyperscalers) are setting up shop in India. Companies like Amazon, Microsoft, and Google are building massive data centre campuses to meet the country’s insatiable demand for cloud services and digital infrastructure.
- Regulatory push for data localisation – India is taking privacy seriously, as evidenced by the enactment of the Digital Personal Data Protection Act of 2023 (DPDPA) This landmark legislation essentially tells companies, “Keep your data close to home.” This policy is driving increased investment in local data centre infrastructure to ensure compliance.
It’s a fascinating convergence of technology, consumer behaviour, and regulation that’s reshaping our digital landscape and creating significant investment opportunities in the process.
Why is India becoming the Silicon Valley of Data Centres?
Let me give you some numbers to show just how fast things are moving. Back in 2019, India’s data centre capacity was about 350 MW. By 2026, we’re looking at hitting 1,645 MW. That’s nearly five times more in just seven years or a 24% compounded annual growth rate (CAGR)!
And where there’s growth, there’s money. Investors are growing cognizant of this fact as well. About USD 5.7 billion is expected to pour into India’s data centre sector by 2026.
Where’s all this happening?
Well, if data centres were cricket teams, Mumbai and Chennai would be at the top of the league table. These two cities alone account for a whopping 81% of all new data centre capacity.
But why are Mumbai and Chennai attracting the lion’s share of the expansion?
It’s because they have some key advantages. They’re coastal, which means they’re close to the undersea cables that connect India to the rest of the digital world. Plus, they’ve already got a strong tech presence.
Delhi NCR, Hyderabad, and Bengaluru too are making moves in the data centre game. These cities have thriving tech scenes and growing digital economies.
And it’s not just Indian companies getting in on the action. Some big global names are teaming up with local players. Yotta, Tata, and Reliance have all shaken hands with NVIDIA, a company that’s leading the AI computing race. These partnerships are bringing cutting-edge AI tech to India, further cementing our place on the global tech stage.
What is the cost of building a Data Centre?
Now, let’s talk about what it takes to actually build one of these digital powerhouses. Spoiler alert – it’s not cheap.
Here’s how the cost split generally is when setting up a greenfield* data centre:
![](https://www.dezerv.in/blog/wp-content/uploads/2024/08/image-4.jpeg)
* A greenfield deployment represents a unique opportunity to build a data centre or network infrastructure from the ground up. Without the constraints or complications presented by legacy systems, you get a blank canvas to design an environment optimised for your needs and future growth.
Setting up a top-notch data centre can cost hundreds of millions of dollars. But the costs don’t stop once the centre is up and running.
About 65% of a data centre’s running costs go towards energy. And of all the power they use, 75% is gobbled up by the IT equipment and keeping it cool.
In the data centre world, costs are often viewed from two perspectives – square footage or computing power. Here, we’re considering the square footage aspect of it:
![](https://www.dezerv.in/blog/wp-content/uploads/2024/08/image-6.jpeg)
To put that into context, consider a massive data centre project in Northern Virginia, often referred to as the data centre capital of the world. A facility spanning 700,000 square feet with 60 megawatts of IT power would require an investment ranging from USD 420 million to USD 770 million. In INR, that translates to an outlay of a whopping INR 3,500-6,500 crore.
If we were to look at it from the computing power angle – experts say that the cost per MW of setting up data centres has reached INR 60-70 crore per MW. This means that a 2,000 MW addition would set us back by INR 12,000-14,000 crore!
These numbers, although approximate, illustrate the immense scale of capital flowing into digital infrastructure. It’s a powerful reminder of the substantial financial commitments underpinning the technologies that are reshaping our world and economy.
As AI becomes more prevalent, the energy demands will only increase. This is a bit of a double-edged sword – we need more computing power to drive AI advancements, but that very computing power is an energy glutton. It’s a challenge, sure, but it’s also an opportunity for innovation.
Challenges and opportunities in the AI era
As we dive deeper into the AI era, data centres are facing some unique challenges. But, with great challenges come great opportunities.
First up, power and cooling. AI workloads, especially those fancy deep learning algorithms, need specialised hardware, like GPUs, that generate enough heat to warm a room! This means we need to get creative with cooling solutions.
But it’s not just about keeping things cool. We also need to think about energy efficiency. This is where things get exciting. Many data centre operators are exploring AI-powered management systems. It’s like using AI to manage AI. These systems can optimise energy usage in real time, cutting waste and improving efficiency.
Now, let’s talk about the carbon footprint. With climate change having taken centre stage in almost all deliberations, data centre operators are under pressure to go green. Many are turning to renewable energy sources like solar and wind power. Some are even getting creative, using the heat generated by data centres to warm nearby buildings. Talk about turning a problem into a solution!
Despite these challenges, India’s potential as a global data hub is enormous. We’ve got a talented IT workforce, a booming digital economy, and a government that’s keen to support tech growth.
Companies that can rapidly scale their AI capabilities by tapping into powerful cloud infrastructure will be best positioned to unlock new innovations and business opportunities.
In summary
As we wrap up, let’s take a moment to reflect. We’re standing at the threshold of an AI-driven future, and data centres are the foundation upon which this future will be built. India, with its rapidly growing digital economy and ambitious plans for AI development, is positioning itself as a global leader in this space.
As AI and cloud computing continue to drive the growth of data centres, we can draw an interesting parallel to the warehouse boom that occurred in India in recent years.
When e-commerce giants like Amazon and Flipkart rapidly expanded their operations in India, they needed to build out vast warehousing capabilities to support faster deliveries to consumers. Piggybacking on the e-commerce boom, the total warehousing space in India grew from 111 million sq ft to 251 million sq ft between 2016 and 2021, a 126% increase.
Similarly, the explosion of data centres is not only laying the groundwork for the widespread adoption of AI and cloud services but also hinting at potential opportunities in the real estate sector.
So, whether you’re a business leader, an investor, or just someone interested in where technology is heading, keep your eye on this space. The data centre boom and the AI revolution aren’t just changing the tech landscape – they’re poised to reshape our entire economy. And India? We’re not just along for the ride – we’re in the driver’s seat.
Disclaimer:
The information contained in this document is for informational purposes only.