Inside Deloitte’s Global AI Data Centre Hub

Deloitte is establishing a global AI Infrastructure Centre of Excellence as companies struggle to build the specialised data centres required for AI workloads.

The professional services firm is responding to mounting client demand for infrastructure that can handle AI’s unique requirements. 

Unlike traditional computing, AI applications need vastly more processing power, generate significantly more heat and require different security protocols.

“As AI adoption accelerates across industries globally, organisations are demanding infrastructure that they can scale with speed, security and efficiency,” says Heather Stockton, Deloitte Global Consulting Services, Technology and Transformation leader.

“Our AI Infrastructure CoE enables clients to move from experimentation to enterprise-grade AI with confidence, while enabling performance, resilience and long-term scalability.”

Deloitte’s data centre strategy 

The Centre forms part of Deloitte’s Silicon2Service offering, which takes clients from foundational silicon components through to business-ready AI solutions. 

Rather than piecemeal fixes, Deloitte provides end-to-end capabilities spanning strategy, design, implementation and operations.

Many organisations have discovered that scaling successful AI pilots into production systems requires entirely different infrastructure approaches. 

Early experiments often rely on cloud services, but enterprise deployments demand more sophisticated planning around capacity, security and regulatory compliance across multiple jurisdictions.

How GPU technology drives specialised cooling requirements

The Centre focuses heavily on next-generation graphics processing units (GPU) technologies, which have become essential for AI workloads. 

GPUs can perform multiple calculations simultaneously, making them ideal for machine learning (ML) tasks. 

However, these processors generate substantially more heat than conventional chips found in traditional data centres.

This has driven Deloitte to invest in advanced cooling systems alongside networking and storage solutions designed specifically for high-performance AI workloads. 

The infrastructure would overwhelm standard data centre setups that weren’t built with AI in mind.

“AI is reshaping industries and redefining competitive advantage and the infrastructure that powers it must be equally transformative,” says Romal Shetty, CEO of Deloitte South Asia. 

“With the launch of our global AI Infrastructure CoE, we are not only investing in advanced capabilities but also laying the foundation for the future of intelligent enterprise infrastructure.”

Furthermore, Deloitte’s security protocols extend beyond hardware to include cybersecurity frameworks with third-party integrations. 

These prove particularly important as AI workloads often process sensitive data across different countries with varying regulatory requirements.

The power of digital twins

Data centre construction costs have risen 40% since 2020, making design errors increasingly expensive. 

As a result, Deloitte’s Centre employs digital twin technology to help clients visualise and test different configurations before physical deployment.

Digital twins create virtual replicas of physical systems that allow simulation and testing without costly mistakes. 

The technology extends into ongoing operations through real-time monitoring that provides insights for continuous improvement. 

These feed into Deloitte’s AI-driven optimisation methodologies, which aim to reduce operational costs while maintaining performance levels.

The firm has also made substantial investments in technical capabilities, infrastructure and AI talent to support the Centre’s operations. 

It has built specialised infrastructure while expanding its team of data centre professionals.

“What sets this CoE apart is its focus on translating infrastructure complexity into tangible business value,” says Ranjit Bawa, Deloitte Global Technology and Ecosystems & Alliances leader.

“This CoE equips Deloitte’s clients to anticipate change, adapt faster and scale smarter. Keeping pace with AI is just the beginning; the focus is on defining what comes next.”