Jio Eyes Satellite Communications and a Sovereign AI Backbone

Context

Reliance Jio is evaluating the development of a sovereign Low Earth Orbit (LEO) satellite constellation for India. The initiative aims to provide reliable internet connectivity to:

Remote villages
Island communities
Border outposts
Other regions where terrestrial telecom networks face limitations

The proposal is part of a broader effort to strengthen India’s digital sovereignty and reduce dependence on foreign satellite communication infrastructure.

What is a Low Earth Orbit (LEO) Satellite Constellation?

A LEO satellite constellation is a network of satellites operating at altitudes between:

160 km – 2,000 km above Earth

These satellites work together to provide:

High-speed internet
Low-latency communication
Wide geographic coverage

Advantages of LEO Satellites

Lower latency than traditional satellites
Faster internet speeds
Better suitability for real-time applications
Improved coverage in remote and inaccessible areas

Jio’s Satellite Communication Strategy

1. Building Indigenous Satellite Capability

The long-term goal is to establish an Indian-owned satellite communication ecosystem that supports:

Digital inclusion
Strategic autonomy
National security requirements

2. Partnering with Global Operators

Until its own constellation is operational, Jio plans to:

Lease satellite capacity from global satellite operators
Accelerate commercial service rollout
Expand broadband access rapidly

Global Competitors

The move comes amid growing competition from:

Both companies are actively pursuing satellite internet opportunities in India.

3. Development of Ground Station Infrastructure

Jio is investing in ground stations across India.

What are Ground Stations?

Ground stations are terrestrial communication hubs that:

Send signals to satellites
Receive signals from satellites
Connect satellite networks with terrestrial internet infrastructure

Importance

They enable:

End-to-end satellite broadband services
Network management
Data transmission and routing

Sovereign AI Backbone: Reliance’s Next Big Initiative

Jio’s parent company, Reliance Industries Limited, is simultaneously building a Sovereign AI Backbone through its subsidiary Reliance Intelligence.

Location

The AI infrastructure is being developed at:

Jamnagar

and will be powered entirely by renewable energy generated from projects in:

Kutch

Purpose

The project aims to address:

Shortage of AI computing infrastructure
High cost of computing power
Growing demand for AI training and inference

Project Capacity

Phase-I

Capacity: 120 MW
Expected commissioning: End of 2026

GPU Deployment

Reliance plans to deploy advanced:

NVIDIA GB300 GPUs

whose AI inference capability is equivalent to:

More than 75,000 NVIDIA H100 GPUs

during the initial phase.

Future Scale

Once the full 120 MW facility becomes operational:

AI computing capability may exceed
2 lakh H100-equivalent GPUs

making it one of the world’s largest AI computing facilities.

What is Satellite-Based Communication?

A communication satellite is an artificial satellite placed in orbit to transmit and receive communication signals between locations on Earth.

Satellite communication enables:

Voice services
Internet connectivity
Television broadcasting
Defence and military communications

particularly in regions where terrestrial connectivity is unavailable or unreliable.

Types of Communication Satellites

Based on orbital location, communication satellites are classified into four categories:

Orbit	Approximate Altitude	Key Feature
GEO (Geostationary Earth Orbit)	~35,786 km	Appears stationary over Earth
MEO (Medium Earth Orbit)	2,000–35,786 km	Navigation and communication
LEO (Low Earth Orbit)	160–2,000 km	Low latency internet
HEO (Highly Elliptical Orbit)	Highly elongated orbit	Better coverage of polar regions

How Does Satellite Communication Work?

Satellite communication involves three major stages:

1. Uplink

Signal is transmitted from a ground station to a satellite.

↓

2. Transponder

The satellite:

Receives the signal
Amplifies it
Changes its frequency
Relays it back

↓

3. Downlink

The processed signal is transmitted back to Earth stations or end users.

Example

In live television broadcasting:

Broadcaster → Satellite (Uplink)
Satellite processes signal (Transponder)
Satellite → TV stations/Viewers (Downlink)

Current Status of Satellite Communication in India

Although the technology is ready, commercial satellite broadband services are not yet fully operational because:

Satellite spectrum allocation is pending
Regulatory approvals are still under process

Recent Development

Jio Platforms has received approval from:

IN-SPACe

to deploy satellites for gigabit broadband services.

However, additional approvals from the Department of Telecommunications (DoT) are required before operations can commence.

What is a Graphics Processing Unit (GPU)?

A Graphics Processing Unit (GPU) is a specialised processor designed to perform millions of calculations simultaneously.

Unlike a CPU, which handles a smaller number of complex tasks, GPUs excel at:

Parallel processing
Large-scale computations
Graphics rendering
Artificial Intelligence workloads

GPU vs CPU

Feature	CPU	GPU
Design Purpose	General computing	Parallel computation
Core Structure	Few powerful cores	Thousands of smaller cores
Best For	Complex decision-making	Repetitive large-scale calculations
AI Workloads	Limited	Highly efficient

How Does a GPU Work?

To render graphics, GPUs follow a four-stage rendering pipeline:

1. Vertex Processing

Determines position of objects on screen
Applies rotation and perspective calculations

2. Rasterisation

Converts geometric shapes into pixels

3. Pixel/Fragment Shading

Calculates:

Colours
Textures
Shadows
Reflections
Lighting effects

4. Frame Buffer Writing

Final image is stored in memory and sent to the display.

Shaders and VRAM

Shaders

Small programs that perform calculations on many pixels or vertices simultaneously.

Functions

Texture mapping
Lighting calculations
Shadow generation
Reflection rendering

VRAM (Video RAM)

Dedicated memory used by GPUs to:

Store textures
Store 3D models
Process images rapidly

High-bandwidth VRAM enables massive data movement required for AI and graphics applications.

GPU Architecture

GPUs may exist as:

Discrete GPUs

Separate chips on graphics cards
Surrounded by VRAM
Common in gaming PCs and AI servers

Integrated GPUs

Embedded within the CPU/System-on-Chip
Common in laptops and smartphones

Global GPU Market

Dominant Players

NVIDIA’s Market Position

Roughly 90% share of the discrete GPU market.
Strong dominance in data centres due to:
- Advanced hardware
- CUDA software ecosystem

Significance for India

Jio’s proposed satellite constellation and Reliance’s AI infrastructure represent two strategic pillars of India’s technological future:

Satellite Communications

Universal broadband access
Improved connectivity in remote areas
Enhanced digital sovereignty

Sovereign AI Backbone

Indigenous AI infrastructure
Reduced dependence on foreign computing resources
Support for India’s AI ecosystem and digital economy

Together, these initiatives can strengthen India’s position in the emerging domains of space-based communications, artificial intelligence, digital sovereignty, and strategic technology infrastructure.