EnterpriseDB platform cuts datacentre emissions by up to 87%
EnterpriseDB has published independent research showing how its data and AI platform can help large enterprise datacentres reduce emissions by up to 87% and energy consumption by up to 81%.
The study analysed operations at three global financial services companies, all using EDB's sovereign data and AI platform which is built on the open-source Postgres system. According to the results, these organisations have been able to cut emissions and energy consumption by over half on average, while still maintaining the performance required for their business-critical workloads.
Reducing emissions and costs
The report highlights the energy demands that come with managing AI workloads and the impact these are having on electricity consumption within datacentres. According to the research, electricity costs already represent 46% of total data centre spending for enterprises, and power demand is expected to rise at a compound annual growth rate of 45% through to 2027, as predicted by IDC analysis.
In reviewing the deployments, it was found that one institution reduced its number of datacentre core servers from 420,000 to 80,000. This 81% reduction in server hardware drove an associated 87% decrease in carbon emissions, while maintaining required performance levels. Overall, emissions decreased by more than 50% on average across the three companies, with another recording a 94% emissions reduction for its Tier 1 applications.
"Facing these rising costs and concerns, 83% of enterprises globally now rank power efficiency among the top three drivers for rethinking data center architectures in the age of agentic AI, according to EDB research," said Kevin Dallas, CEO at EDB. "The leading 13% of AI-driven enterprises already know this. They're twice as likely to design for energy-efficient model execution. They're proving it's possible to optimize energy use without sacrificing performance, agility, or scale."
Dallas went on to add: "These aren't theoretical gains. It's the operational reality now for the world's largest financial services companies and pinpoints the extraordinary path forward: You can achieve dramatic reductions in data center emissions and energy with no trade-offs in performance or availability. And in many cases, this comes with dramatic cost reductions as well."
Technology for efficiency
The EDB PG AI platform achieves its results by ensuring that computing resources are only used when necessary, based on elastic infrastructure that can scale up or down according to the workload. This approach is paired with various features targeting the needs of modern AI and data estates, including intelligent workload optimisation, AI-driven recommendations, and on-demand serving of AI models to reduce underutilised infrastructure.
Other features include automated management of AI readiness, reducing the need for inefficient manual interventions, and separation of compute from storage to allow independent scaling and further reduction in energy use by offloading less active data.
Hardware is also a contributing factor. EDB's platform is deployed on Supermicro Hyper servers, noted for power efficiency, which are available in modular, liquid-cooled, and GPU-optional environments based on the needs of the customer. Relationships with hardware companies like Supermicro are an important part of the energy and cost savings that can be achieved.
Introducing the Efficiency Calculator
To help organisations evaluate the impact of optimisation on their own infrastructure, EDB has introduced the EDB Postgres AI Efficiency Calculator. The interactive tool provides an instant analysis of how optimising their data platform could reduce operational costs, power usage, and carbon emissions. The recommendations delivered are based on real usage data, offering paths to immediate improvements in efficiency and sustainability outcomes.
The growing focus on AI and digital transformation among enterprises means demand for datacentre resources will continue to climb. Research commissioned by EDB has shown that more than 95% of organisations wish to become sovereign over their own AI and data operations within three years, signalling a potential rise in energy requirements that must be managed. The company's latest research indicates that with attention to platform architecture, hardware, and automated management, significant environmental and financial savings are attainable.
Dallas commented, "The future of AI infrastructure isn't one size fits all. It's modular, power-aware, and built for choice. As energy and performance requirements evolve, enterprises need flexible systems that scale without compromise. That's what we're delivering with EDB Postgres AI."
