The Backbone of AI: Why Data Centres Are Essential for AI Workloads

Artificial intelligence (AI) has captured the imagination of industries worldwide, transforming everything from healthcare and finance to manufacturing and mobility. In just a few short years, AI has gone from buzzword to boardroom imperative, with decision-makers across sectors now prioritising AI adoption to future-proof their businesses. In Asia-Pacific, this transformation is especially pronounced: IDC projects that investment in AI and generative AI technologies in the region will reach US$110 billion by 2028.

One in four businesses that have scaled up their generative AI (gen AI) use are finding greater returns on investments, according to consulting firm BCG. In Asia-Pacific (AP), mature companies are experiencing 25% shorter time-to-market with gen AI products, it reports. Across industries, AI is now top of mind for leaders planning for the future . 

The impact of AI is tangible. In healthcare, AI enables faster interpretation of MRI scans, leading to better patient care. In mobility, self-driving vehicles, may become more common on the streets in the years ahead, and in manufacturing, AI-driven analytics are helping reduce material wastage and promote more sustainable production. For example, manufacturers are using AI for better analysis of production yields to reduce material wastage and optimise output.

However, the success of any AI use case relies on robust data processing capabilities. AI applications demand vast computing power, storage, and networking infrastructure to function effectively.

Data centres are the essential foundation powering  artificial intelligence. While much of the attention is on the impressive capabilities of AI applications, it is purpose-built, resilient data centre infrastructure that enables these innovations to thrive. At STT GDC, we  deliver the reliable, high-performance digital infrastructure that underpins AI’s progress and supports our customers’ evolving needs.

The foundation for AI advancement
AI workloads are unlike anything traditional IT infrastructure was designed to handle. Training a single advanced AI model can require processing petabytes of data and running trillions of calculations — a computational feat that would overwhelm legacy, on-premises server rooms. Next-generation data centres, engineered for high-density computing, robust networking, and advanced cooling, have become the essential foundation for AI innovation.

AI-ready data centres are not just larger versions of their predecessors. They are purpose-built environments designed to support both the training and inference phases of AI. Training involves feeding vast datasets into deep learning algorithms, demanding massive parallel processing, high-density GPU clusters, and extensive storage. Inference, by contrast, requires ultra-low latency and high-speed networking to deliver real-time AI-powered insights and responses — think of an AI chatbot generating answers or an edge device rendering images instantly for a user.

The New Standard: Purpose-Built for AI
The rise of AI is fundamentally reshaping data centre design. Today’s facilities can support rack densities of 12kW to 15kW for general cloud computing, but with some AI workloads pushing requirements to 30-50kW per rack or even a tenfold increase to 150kW per rack, this surge in power density has driven rapid innovation in everything from liquid and immersion cooling systems to advanced fibre-based interconnects and modular, scalable infrastructure.

But the transformation goes further. AI data centres offer:

• Future-proofing, with designs and advanced management tools that allow for upgrades and expansion as AI workloads evolve and grow.

• Sustainability, as operators like STT GDC integrate renewable energy, efficient liquid cooling, and AI-driven optimisation to reduce environmental impact and support long-term digital growth.

• Physical security for sensitive AI models and intellectual property, with advanced access controls and 24/7 monitoring.

Organisations will seek to build up their AI capabilities differently, as the technology matures. Some may rely on hyperscale cloud providers for GPU-as-a-service for the required performance. Others may prefer a colocation data centre, where they can deploy their own compute infrastructure and related equipment to run their AI workloads. 

Why data centres are indispensable for AI
The world’s insatiable appetite for AI is driving an unprecedented surge in demand for data centre capacity. According to McKinsey, global demand for data centre capacity could rise by up to 22% annually through 2030, with as much as 70% of that demand coming from AI-ready facilities. Hyperscalers and cloud providers are racing to build sufficient capacity, but even their efforts may not be enough to avoid a supply crunch.

Traditional IT infrastructure simply cannot keep up. The scale, speed, and complexity of AI workloads require the specialised environments only next-generation data centres can provide. This is why, as AI becomes ubiquitous and business-critical, the data centre is the backbone of the intelligent age.

AI-ready for the future
At STT GDC, we see ourselves not just as data centre operators, but as enablers of the digital future. Our facilities are engineered to meet the unique demands of AI workloads — delivering the high-performance computing, storage, and networking that enterprises, governments, and innovators need to unlock the next wave of AI breakthroughs. Sustainability is woven into our approach, with investments in renewable energy, advanced cooling, and AI-driven efficiency that future-proof both our operations and our customers’ ambitions.

Reinforcing this commitment to provide robust, scalable infrastructure that supports our customers’ deployment of advanced AI platforms and high-density computing, STT GDC’s data centre facilities — STT Singapore 6 and STT Bangkok 1 —  have achieved certification under the NVIDIA DGX-Ready Data Center program. This milestone enables us to offer customers access to world-class, state-of-the-art infrastructure purpose-built for AI workloads, empowering enterprises in Southeast Asia to simplify and accelerate their AI initiatives.

As AI continues to reshape industries and societies, one thing is clear: the success of tomorrow’s AI innovations will depend on the strength, resilience, and intelligence of the data centres that power them. At STT GDC, we are proud to be the backbone of this transformation — enabling our customers to imagine, build, and scale the future of AI.