TCS HyperVault’s ₹70,000 Crore AI Data Centre in Hyderabad: What It Means for India’s AI Infrastructure

Tata Consultancy Services (TCS) is making one of its biggest bets yet on the physical infrastructure behind artificial intelligence.

Through its AI infrastructure subsidiary HyperVault AI Data Centre Ltd, TCS has secured 264 acres of land in Hyderabad to develop an AI-focused data centre campus with capacity of up to 1 gigawatt (GW). HyperVault and its partners are expected to invest up to ₹70,000 crore to build and manage the infrastructure.

The project is being designed specifically for frontier AI companies, hyperscalers and global enterprises that need high-density computing infrastructure for AI training, inference and other advanced workloads.

But the headline ₹70,000 crore figure is only one part of the story.

The more important development is that TCS is attempting to move deeper into the AI infrastructure stack — from data centres and compute to cloud, engineering, enterprise transformation and AI services.

TCS calls this strategy “Infrastructure-to-Intelligence.”

And Hyderabad could become one of the physical foundations of that strategy.

TCS HyperVault Hyderabad AI Data Centre: Key Facts

These details come from TCS HyperVault's announcement and have been independently reported by multiple Indian business publications.

What Is TCS HyperVault?

HyperVault is TCS's dedicated AI infrastructure business, established with strategic investment partner TPG.

The idea behind HyperVault is different from simply building another conventional data centre.

Traditional data centres support a wide range of enterprise workloads such as websites, databases, cloud applications and business software. AI workloads can require dramatically higher computing density, faster networking and much more sophisticated cooling.

HyperVault is being designed specifically around those requirements.

TCS says the business is intended to provide secure, scalable and AI-ready data centre infrastructure for organisations that need to train, run and scale AI systems. Its current plans call for more than 1GW of new AI-ready capacity across key locations in India.

This puts HyperVault at the infrastructure layer of TCS's broader AI strategy.

Why Is TCS Building a 1GW AI Data Centre?

The answer is simple: AI requires enormous amounts of computing power.

Training and running large AI models requires specialised processors, particularly GPUs and other AI accelerators. As AI adoption moves from experimentation to production, companies need infrastructure capable of running these systems continuously.

The requirements are very different from those of a typical corporate data centre.

• High-density GPU clusters
• High-speed networking
• Large and reliable power supply
• Advanced cooling
• High availability
• Low-latency connectivity
• Strong security
• Data localisation capabilities

This is why the TCS HyperVault project is being built around high-density, liquid-cooled computing infrastructure.

The company says the facility will support high-density GPU deployments for AI training, inference and advanced computing workloads.

What Does 1GW Mean for an AI Data Centre?

One gigawatt equals 1,000 megawatts of power capacity.

However, it is important not to interpret the announcement as meaning that the entire campus will immediately consume 1GW of electricity.

The 1GW figure represents the planned scale of the data centre campus as it is developed.

The project will be built in phases, with the pace of expansion determined by customer demand and changes in AI technology.

This phased model is particularly important because AI infrastructure is changing rapidly. GPU performance, rack densities, networking requirements and cooling technology are all evolving.

Building the entire 1GW capacity at once would therefore create unnecessary capital and technology risk.

Why Is the Investment Up to ₹70,000 Crore?

The ₹70,000 crore figure has attracted significant attention because it makes the HyperVault Hyderabad project one of the largest proposed AI infrastructure investments in India.

But the wording matters.

It is an investment potential of up to ₹70,000 crore, not an immediate ₹70,000 crore expenditure.

HyperVault and its partners expect to invest the amount over the development and management of the infrastructure, with construction proceeding in phases.

Reuters estimates the potential investment at approximately $7.4 billion.

The capital requirements of a project of this scale extend far beyond the physical data centre building.

Investment can include:

• Land development
• Data centre buildings
• Electrical infrastructure
• Power distribution
• Backup power systems
• Cooling systems
• GPU and computing infrastructure
• Networking equipment
• Fibre connectivity
• Security systems
• Operations infrastructure

As a result, the eventual investment will depend heavily on how quickly capacity is contracted and what type of computing infrastructure customers require.

The OpenAI Connection Makes HyperVault More Important

One of the most significant developments surrounding TCS HyperVault is its relationship with OpenAI.

OpenAI has identified TCS HyperVault as its first customer-facing infrastructure partner in India under its broader India initiative.

The initial arrangement involves 100MW of capacity, with the potential to scale to 1GW over time.

OpenAI says the infrastructure is intended to support its advanced AI models in India while addressing requirements around data residency, security, compliance and lower latency.

This distinction is important:

• 100MW is the initial OpenAI capacity.
• 1GW is the potential scale of the OpenAI arrangement over time.
• 1GW is also the planned capacity of the Hyderabad HyperVault campus itself.
• The entire Hyderabad campus should not automatically be described as an OpenAI data centre.

Instead, OpenAI provides an important example of the kind of global AI customer HyperVault is attempting to serve.

Why Does OpenAI Need AI Infrastructure in India?

AI companies are increasingly looking beyond simply providing access to models through the public internet.

Large-scale AI deployment creates requirements around:

• Compute capacity
• Latency
• Data residency
• Security
• Regulatory compliance
• Enterprise workloads
• Government workloads

OpenAI says its partnership with Tata Group is intended to build local AI-ready data centre capacity as part of its broader Stargate initiative.

For India, this potentially means more AI compute can be located within the country instead of relying entirely on infrastructure outside India.

Why Hyderabad Was Chosen for the TCS AI Data Centre

Hyderabad has already developed into one of India's major technology centres.

But the requirements of AI infrastructure go beyond software talent.

A large AI data centre requires:

• Large land parcels
• Reliable electricity
• High-quality fibre connectivity
• Cooling infrastructure
• Engineering capabilities
• Technology talent
• Government support
• Data centre ecosystem

The 264-acre HyperVault site gives TCS significant physical room to develop the campus in phases.

Telangana has also positioned Hyderabad aggressively as an AI and technology hub, and the state government has described the HyperVault project as a major development for its technology ecosystem.

Hyderabad Could Become an AI Compute Hub

The significance of the project extends beyond TCS.

Once a large AI data centre ecosystem begins developing around a city, it can attract additional investment into supporting industries.

These can include:

• Power infrastructure
• Renewable energy
• Data centre construction
• Cooling technology
• Networking
• Semiconductor infrastructure
• Cloud services
• AI software
• Cybersecurity
• Data centre operations

This creates the possibility of an AI infrastructure cluster rather than a single isolated data centre.

TCS Is Building More Than a Data Centre

This is perhaps the most important strategic point about HyperVault.

TCS already has a massive global enterprise customer base and operates across consulting, cloud, engineering, software and IT services.

Its challenge in the AI era is to move beyond helping companies experiment with AI and become a company capable of supporting AI across the enterprise technology stack.

HyperVault gives TCS an infrastructure layer to complement those existing capabilities.

The resulting model can look something like this:

Data Centre → Compute → Cloud → AI Models → Enterprise Integration → AI Applications

TCS already operates heavily in the upper layers of this stack.

HyperVault gives it an opportunity to move closer to the infrastructure layer.

What Does “Infrastructure-to-Intelligence” Mean?

TCS describes HyperVault as part of its Infrastructure-to-Intelligence strategy.

The concept is that enterprises do not simply need an AI model. They need the infrastructure, data, applications, security, governance and integration required to make AI useful inside their businesses.

TCS's CEO has said that customers increasingly want partners capable of building and managing a unified control plane across the AI stack — from infrastructure to intelligence.

This is an important change in positioning.

TCS is effectively trying to become an AI infrastructure and enterprise intelligence integrator, rather than competing purely as an IT services provider.

TCS Is Building an AI Ecosystem Around HyperVault

HyperVault is not operating in isolation.

TCS has announced partnerships with companies including OpenAI, AMD and ABB as part of its broader AI infrastructure strategy.

Its partnership with AMD includes work on a rack-scale AI infrastructure design based on AMD's Helios platform. TCS and AMD also plan to offer an AI-ready data centre blueprint supporting up to 200MW of capacity and work with hyperscalers and AI companies on data centre build-outs in India.

That is strategically important because a data centre company doesn't need to manufacture every component itself.

It can build the infrastructure layer while working with semiconductor, networking, cooling, power and cloud partners.

Why Liquid Cooling Matters

AI GPUs generate substantial amounts of heat, particularly when large numbers of processors are packed into a small physical space.

Traditional air cooling becomes increasingly difficult as rack density rises.

HyperVault therefore plans to use direct-to-chip liquid cooling for next-generation GPU and semiconductor workloads.

TCS describes liquid cooling, high rack density, resilient power and fibre connectivity as key elements of its AI-ready data centre architecture.

This is one reason an AI data centre is not simply a normal data centre with more GPUs installed.

The entire physical infrastructure has to be designed around the computing workload.

Power Could Become the Biggest Constraint in AI

As AI computing expands, electricity is becoming a strategic input for the technology industry.

A large AI campus requires reliable power around the clock.

It also needs infrastructure capable of delivering that power at extremely high density while maintaining uptime.

This means India's AI race is increasingly becoming a race involving:

• Electricity
• Renewable energy
• Transmission infrastructure
• Data centres
• Cooling technology
• GPU supply
• High-speed networking

TCS says HyperVault will incorporate green-energy principles and evolve toward fully green power, while also focusing on water-neutral design.

What Will the 1GW Hyderabad Campus Be Used For?

The campus is designed to support several categories of AI workloads.

AI Model Training

Training large AI models requires enormous amounts of computing capacity and high-speed communication between GPUs.

AI Inference

Once an AI model is deployed, users and businesses continuously send requests to it. Running those requests at scale is called inference.

Enterprise AI

Large companies increasingly want AI systems embedded into customer service, software development, finance, supply chain, cybersecurity and other business functions.

Advanced Computing

The infrastructure can also support other computationally intensive workloads that require high-performance computing.

How Could TCS Make Money From HyperVault?

The business model is different from traditional TCS software services.

At the infrastructure level, revenue can potentially come from providing customers with access to data centre capacity, compute infrastructure and associated services.

Over time, TCS can potentially combine this with its existing enterprise capabilities.

For example:

Customer needs AI infrastructure

↓

HyperVault provides the data centre capacity

↓

TCS provides cloud and engineering integration

↓

TCS deploys AI applications into the customer's business

That creates an opportunity for TCS to capture value at multiple layers of the AI adoption cycle.

How Big Is TCS's AI Business Already?

The HyperVault investment is also happening alongside rapid growth in TCS's existing AI-related business.

TCS said its annualised AI revenues surpassed $2.3 billion in Q4 FY26, reflecting accelerated deployment of AI solutions. The company also said its HyperVault investment helped strengthen strategic partnerships with OpenAI, AMD and ABB.

This is important context.

TCS isn't building HyperVault before having an AI business. It is building infrastructure while its existing AI and enterprise business is already expanding.

How Many Jobs Could the TCS Hyderabad AI Data Centre Create?

TCS and its partners expect the project to generate several thousand direct and indirect jobs.

The employment impact can extend well beyond data centre employees.

A project of this scale can create demand for:

• Electrical engineers
• Mechanical engineers
• Cooling specialists
• Network engineers
• Data centre technicians
• Construction workers
• Security professionals
• Cloud engineers
• AI infrastructure specialists
• Operations teams

Telangana's chief minister has separately urged HyperVault to target an inauguration by June 2, 2028, although the company's announced development model remains phased according to demand and technology requirements.

What Does This Mean for India's AI Infrastructure?

India has traditionally been known globally for its software engineers, IT services and outsourcing industry.

The next stage of the AI economy requires something different.

It requires the physical infrastructure to actually run AI.

That includes:

• GPUs
• AI data centres
• Cloud infrastructure
• Power
• Networking
• Cooling
• Storage
• Security

The TCS HyperVault Hyderabad project is therefore significant because it represents India's transition from being primarily a consumer and developer of AI services toward becoming a location where significant AI computing infrastructure is physically deployed.

Is ₹70,000 Crore Really Going Into the Project?

This is one of the most important questions surrounding the announcement.

Not immediately.

The announced figure is up to ₹70,000 crore for HyperVault and its partners to build and manage the infrastructure.

The campus will be developed in phases. The actual capital deployment will therefore depend on customer demand, capacity commitments, technology requirements and the speed at which the project is built.

Readers should therefore avoid interpreting the headline as “TCS has already spent ₹70,000 crore on a data centre.”

It is a long-term infrastructure investment opportunity with a potential total outlay of that scale.

What Are the Risks?

The opportunity is enormous, but the project is not without risks.

AI Infrastructure Is Capital Intensive

Data centres require large amounts of upfront capital, unlike software products that can scale with relatively little physical infrastructure.

Technology Changes Quickly

AI accelerators and data centre architectures are evolving rapidly. Infrastructure must remain adaptable to future generations of GPUs and AI systems.

Power Availability

Large AI campuses require enormous amounts of reliable electricity. Power availability can become a constraint on expansion.

Customer Concentration

Large AI data centres require major customers to justify capacity expansion. Long-term customer commitments therefore become critical.

GPU Economics

The economics of AI infrastructure depend partly on GPU utilisation and the commercial value generated by the compute capacity.

The phased development approach helps HyperVault manage some of these risks by aligning capacity expansion with actual customer demand.

TCS HyperVault vs Traditional Data Centres

Why the TCS HyperVault Project Matters for Investors

From an investor's perspective, the project is interesting because it represents a potential new growth engine around an existing IT services company.

TCS has historically generated most of its revenue through technology services, consulting and enterprise transformation.

AI infrastructure introduces a different economic model.

Instead of selling only human expertise and software services, TCS can potentially participate in the underlying infrastructure required for AI workloads.

However, investors should also remember that HyperVault is a long-term infrastructure bet. The ₹70,000 crore figure should not be treated as near-term TCS revenue or profit.

The commercial success of the project will ultimately depend on capacity utilisation, customer contracts, capital expenditure, operating costs and returns on the infrastructure deployed.

What Happens Next?

The immediate focus will be on converting the announced project into physical infrastructure.

The key developments to watch will include:

• Construction milestones
• Power and infrastructure approvals
• Customer commitments
• OpenAI capacity deployment
• GPU and semiconductor partnerships
• Additional hyperscaler customers
• Expansion beyond the initial capacity
• Actual capital expenditure
• Commercial commissioning

The company's ability to secure additional customers will be particularly important because the campus is designed to scale according to demand.

Conclusion: TCS Is Betting on the Infrastructure Behind AI

The TCS HyperVault Hyderabad project is much more than a ₹70,000 crore data centre announcement.

It represents a strategic bet on one of the biggest infrastructure requirements created by the AI boom.

TCS is combining its existing strengths in enterprise technology, cloud, engineering and AI with a new infrastructure layer through HyperVault.

The 264-acre Hyderabad campus is planned to reach 1GW of capacity, with up to ₹70,000 crore of investment from HyperVault and its partners.

OpenAI's initial 100MW commitment, with the potential to scale to 1GW, provides an early example of the kind of demand the infrastructure is intended to serve.

The project also shows why the next phase of India's AI story will not be limited to software developers and AI applications.

AI needs electricity. AI needs GPUs. AI needs cooling. AI needs data centres. AI needs networking.

And companies that control those infrastructure layers could capture a significant part of the economic value created by the AI boom.

For TCS, HyperVault is an attempt to ensure that the company is not merely helping businesses use artificial intelligence, but is also building some of the infrastructure required to power it.

Frequently Asked Questions About TCS HyperVault

What is TCS HyperVault?

TCS HyperVault is the AI infrastructure business of Tata Consultancy Services, backed by the wider Tata ecosystem and TPG. It is being developed to provide AI-ready data centre infrastructure for high-performance computing workloads.

Where is the TCS HyperVault AI data centre?

The announced 1GW AI data centre campus is being developed in Hyderabad, Telangana, on a 264-acre site.

How much is TCS investing in the Hyderabad AI data centre?

HyperVault and its partners are expected to invest up to ₹70,000 crore to build and manage the infrastructure. The project will be developed in phases, so this should not be interpreted as an immediate ₹70,000 crore expenditure.

What is the capacity of the TCS Hyderabad AI data centre?

The planned HyperVault campus has capacity of up to 1GW.

Is OpenAI using TCS HyperVault?

Yes. OpenAI has identified TCS HyperVault as its first customer in India, with an initial 100MW capacity arrangement and potential to scale to 1GW over time.

Why is liquid cooling being used?

AI workloads can create very high heat densities because large numbers of GPUs and other accelerators operate together. HyperVault is therefore designed around high-density computing and direct-to-chip liquid cooling.

When will the TCS HyperVault Hyderabad data centre be operational?

The campus is being developed in phases. Telangana's chief minister has publicly urged HyperVault to target an inauguration by June 2, 2028, but the broader project will be rolled out according to customer demand and technology requirements.

Why is TCS building AI data centres?

TCS is using HyperVault to expand its AI strategy from enterprise AI services into the underlying infrastructure needed to train, deploy and scale AI systems. The company describes this broader strategy as “Infrastructure-to-Intelligence.”