Electricity demand from data centres has surged in recent years and is expected to climb further as AI workloads expand, while rising energy costs and environmental pressures make traditional expansion through larger facilities and more hardware increasingly unsustainable. Three industry experts discuss how operators can scale infrastructure while managing energy constraints and costs through efficiency-focused design, renewable energy integration and workload optimisation.
Global data centre electricity demand has already seen a huge shift in recent years and is only expected to surge higher over the next decade as AI workloads grow, forcing operators to rethink how infrastructure expands.
Traditional expansion models centred on larger facilities and more hardware are no longer sustainable as energy costs rise, and environmental pressures intensify. Therefore, the question now is: How can data centre operators scale while managing energy constraints, sustainability pressures and rising costs?
Walid Issa, Senior Manager, Solutions Engineering – Middle East & Africa, NetApp; Jonathan Knepher, VP, Site Reliability Engineering, Forcepoint; and Alfred Manasseh, Co-Founder & COO at Shaffra, discuss this and explain how efficiency-driven design, renewable energy integration and workload optimisation can help meet rising performance needs while controlling power consumption and costs.
Walid Issa, Senior Manager, Solutions Engineering – Middle East & Africa, NetApp:
Data centre operators are expanding capacity at a time when energy has become the limiting factor. According to the International Energy Agency, electricity demand from data centres is expected to rise sharply through 2030, driven largely by AI workloads. In the Middle East, projections show regional capacity could grow from roughly 1.2GW to 3.3GW within five years, more than doubling as digital infrastructure investment accelerates.
Scaling data centre capacity in today’s environment requires a fundamental shift toward smarter, more efficient infrastructure. Data centre operators can no longer rely on expanding physical footprint alone; instead, they must optimise how data is stored, moved and consumed.
The first lever is efficiency. Modern, AI‑enabled storage platforms like NetApp help reduce power consumption by placing the right data on the right tier at the right time. Intelligent tiering, compression and deduplication can reduce the effective capacity footprint by up to 60–70%, meaning more performance is achieved with fewer watts and less space.
Hybrid cloud approach is another option. By extending workloads to hyperscalers or local cloud providers, DC operators can burst capacity without expanding on‑premises energy usage.
Finally, observability and automation are essential. With real‑time insights into energy use, cooling efficiency and workload behaviour, operators can automatically balance resources, eliminate inefficiencies and delay capex-heavy expansions. Combining this with renewable‑ready designs and liquid‑cooling‑friendly layouts positions data centres for long-term resilience.
In summary, sustainable scale isn’t about building bigger, it’s about building smarter. With efficient storage, hybrid cloud flexibility and intelligent automation, operators can grow capacity responsibly while meeting rising performance demands.
Jonathan Knepher, VP, Site Reliability Engineering, Forcepoint:
Data centre expansion can no longer focus solely on capacity – it must be tied to resilience, sustainability and long-term operational stability. While network capacity keeps scaling, energy generation and distribution infrastructure still face physical and regulatory limits. Power availability is quickly becoming one of the main constraints on growth.
To address this, here is what operators need to prioritise:
- Generate as much local renewable energy as possible. Producing most power on-site or adjacent reduces grid dependence and transmission losses. Integrating local power storage in a grid-interactive model can also support surrounding communities by contributing capacity during peak periods.
- Scale beyond a single facility. Architecting information systems across multiple data centres ensures workloads can move dynamically if one site hits a power constraint or needs to shed load. Asynchronous workloads should run only when excess resources are available, maintaining efficiency without compromising performance.
- Apply proven green data centre practices. Using cold air economisers, efficient cooling strategies and high ceilings for thermal headroom, along with designs that allow for balanced peak power usage, all help reduce overall energy demand.
- Integrate local communities. As AI becomes more adept in entry-level roles across industries, it’s important to ensure that the remaining operational and early-career technical roles are sourced locally where these facilities exist. Keeping people-first expertise close to the infrastructure strengthens long-term economic impact and creates shared ownership.
Building in this distributed manner also reinforces cybersecurity resilience. Systems designed to operate across multiple environments can absorb disruptions at one site, while embedding distribution into normal operations ensures continuity and supports sustainable growth.
Alfred Manasseh, Co-Founder & COO at Shaffra:
Scaling data centre capacity sustainably today requires moving beyond simply adding more infrastructure, especially as forecasts suggest that the global power demand from data centres could rise by about 165% by 2030 compared with 2023. While much of the current focus remains on expanding racks and power, long-term success depends far more on how efficiently computing resources are used.
Much of today’s demand is still driven by short-term AI model training, which is expensive and energy-intensive. In contrast, everyday enterprise applications such as workflow automation, digital employees and AI-powered operations create steady, predictable demand that supports long-term utilisation.
Stronger partnerships with telecommunications providers can also play an important role in scaling more effectively. As strategic partners or co-owners, telcos bring long-term power agreements, nationwide fibre networks, regulatory expertise and established enterprise customers. This makes it easier for operators to optimise energy use, connect edge and core infrastructure, reduce latency and support national digital objectives.
At the same time, energy challenges call for smarter use of computing resources rather than simply increasing capacity. Techniques such as model optimisation, quantisation and distillation, alongside edge deployment, can significantly reduce unnecessary processing. When AI systems are designed around task efficiency and real workflow needs, they naturally consume less energy and operate at lower cost.
Sustainability must also make financial sense. With data centre power demand expected to grow sharply over the next decade, operators cannot afford to scale ahead of real adoption. Infrastructure must grow in line with enterprise usage to avoid pressure on margins. Data centres are most resilient when supported by sectors deploying AI at scale, including government, workforce automation and core business operations.
Looking ahead, successful expansion will be distributed, sovereign, telco-integrated and application-led. Operators that align infrastructure, models and real-world deployment into a single, demand-driven ecosystem will be best positioned for long-term growth.
