How Deutsche Telekom Aligned Performance and Sustainability with Google Cloud and Gemini

By Denise Pearl, Global Market Lead, Sustainability & World Models, Google Cloud

Sustainability doesn’t have to be a business trade-off. When data processing and core operations systems are sustainable by design, they create infrastructure that is faster and more secure while optimizing energy consumption and reducing carbon emissions. To scale AI innovation responsibly, it's important to consider sustainability in core design choices alongside price and performance.

Deutsche Telekom, Europe’s largest telecommunications company, serves as a prime example for how modernizing infrastructure with Google Cloud and AI aligns business objectives with environmental responsibility.

Modernizing Systems

Storing data and running applications “on-premise”—i.e. utilizing servers and hardware physically located within a company's facilities—is often less efficient compared to cloud-based services. To modernize, Deutsche Telekom evaluated its operations through Google Cloud’s 4M "Sustainable by Design” framework:

• Machine: Select the right machine for their workloads.

• Model: Build or select accurate, efficient models for resource efficiency.

• Mechanization: Use intelligent, serverless data, analytics, and AI services to handle massive data volumes with superior efficiency.

• Map: Choose where to host your AI workloads based on regional energy profiles.

Guided by these principles, Deutsche Telekom migrated its SAP platform from on-premise to Google Cloud. SAP centralizes business operations from separate programs into a single database enabling companies to connect their departments and work more efficiently. For Deutsche Telekom, many of these extensive operations, responsible for critical business practices, had fallen behind the latest efficiency standards. Cloud migration modernized and centralized Deutsche Telekom’s SAP applications, creating scalable, efficient infrastructure that is resilient to future application updates. During the migration, their teams balanced a less than 0.5 millisecond data latency requirement with sustainability performance metrics, selecting a data center in the Netherlands for superior energy and water efficiency.

Accelerating Transformation with AI

To modernize its infrastructure, Deutsche Telekom leveraged Gemini during the Google Cloud migration to conduct a comprehensive code audit, identifying unnecessary processes and aligning software with modern protocols. By utilizing AI to test workloads for energy efficiency, the team ensured the new cloud environment was optimized for both speed and sustainability.

“Using Gemini as an archaeologist, it helped us check our code and rework it to modernize our whole system,” said Amir Kangarlou, Senior Executive Program Manager and Corporate Responsibility Manager for the IT group at Deutsche Telekom.

This AI-augmented approach allowed Deutsche Telekom to complete a complex migration, which would typically require 6 to 12 months, in just six weeks, achieving a full transition from on-premises operational systems to the cloud within two years.

Delivering Tangible Energy and Performance Gains

Deutsche Telekom reports that migration to Google Cloud resulted in a 40% reduction in energy use annually to support these SAP workloads. This optimization translates to 3.8 gigawatt-hours of energy saved, enough energy to power over 350 homes for an entire year.¹

Beyond the environmental impact, the migration proved that high-performance computing, specifically meeting the 0.5 millisecond latency goal, is inherently linked to resource efficiency; optimized code requires less computing time and, consequently, less power.

Looking Ahead

Deutsche Telekom continues to leverage Gemini and Google Cloud to maintain sustainable software development practices. By auditing their data processing systems for efficiency at the point of development, the organization has established a model for how global enterprises can scale their digital operations while significantly reducing their environmental footprint.