Electricity lit up the 20th century, powering the Industrial Revolution and illuminating our cities. Now, in the 21st century, it is “compute” — or the raw processing power that they will need to run artificial intelligence. Mukesh Ambani, a titan of industry well accustomed to shaking up businesses with his company Jio, has just gone all in. The bet? Massive one-gigawatt data centre network. Polluting data is coming from Jamnagar, India’s industrial heart. This is not just a server farm; it is the building of a national nervous system that will democratise artificial intelligence for more than a billion people.
Understanding the Scale: What is a Gigawatt?
“One gigawatt” means nothing to the average person. Let’s break it down. One gigawatt is equivalent to the energy needed to power a mid-sized city, or roughly 700,000 homes. To allocate a full third of it for data centre use is a monumental feat and the clearest possible declaration of intent.
Data is the car; data centres are the car factory. When you pose a question to ChatGPT or use an app that translates between dozens of languages, the “thinking” doesn’t happen on your phone. Your phone sends a request to a sprawling, cavernous warehouse of racks full of servers — computers with no screens that instead crunch numbers — and they send the response back. Those servers produce enormous heat and require vast amounts of electricity to operate and stay cool. With the creation of this capacity in Jamnagar, Reliance taps into our own energy -literally- to power the weapon for fighting the wars of tomorrow.
The Edge of the Network
The announcement wasn’t limited to large buildings. Ambani referred to a layer called the “edge compute” layer. Imagine a library. The central data centre is the Library of Congress; they have everything, but you have to go there and wait to check out a book. Edge computing is akin to erecting a little branch library on every corner.
By getting processing power closer to the user (the “edge”), latency, or the time it takes for data to travel, plummets. This is crucial for AI. If you are looking for real-time translation of a conversation, or if an autonomous vehicle must identify a pedestrian, there is no time to waste sending the data across the country and back. Jio’s strategy is to embed this intelligence into a cellular network and make AI real-time.
AI for All: Beyond the Buzzword
“AI for All” is perhaps the crassest marketing copy you can imagine, but there’s infrastructure behind it that enables it to become a reality. Nowadays, high-quality AI tends to be either subscription-based or limited to expensive hardware. The idea, as with the mobile data gambit, is to bring down the cost of “intelligence” by generating immense domestic compute capacity and turning artificial intelligence into a commodity.
We envision a future in which a farmer in an isolated village can run AI-driven data and crop analysis on a low-end smartphone, or a student can use a personalised AI tutor without a high-speed fibre connection. Reliance is essentially building highways and bridges for a digital economy, where intelligence can be as easily accessed as water or gas, coursing freely into every Indian home.
