Making Telcos Run Leaner and AI-Ready with AMD

Oct 20, 2025

Abstract background

Making Telcos Run Leaner and AI-Ready with AMD  

Telecommunications companies are under extraordinary pressure to modernize their infrastructure while navigating a sea of evolving demands. The explosive growth of 5G, the convergence of fixed and wireless networks, the rising need for real-time AI applications, and the push for more sustainable operations have made traditional approaches nonviable

At the same time, telco IT and business leaders face mounting challenges including managing surging data traffic with ultra-low latency, scaling services cost-effectively while reducing power and rack space, and supporting diverse workloads ranging from legacy voice to cloud-native apps and AI. They also need to address aggressive environmental sustainability goals without compromising performance.

Industry leaders like Swisscom and Chunghwa Telecom Information Technology Group (CHT ITG) are transforming their infrastructure for AI by leveraging AMD technology to help them meet these challenges head-on.

Scaling Performance and Productivity

High-performance computing is no longer optional in telecommunications. With growing demands on throughput, latency, and user experience, telcos need to be able to process more data in less time, at every level of the network. To meet these escalating demands, some leading telecom operators are turning to advanced server architectures to boost efficiency and performance, resulting in gains in density, capacity, and energy savings.

Swisscom nearly doubled the virtual CPU density per server in its telco cloud, unlocking greater capacity for both telecom and third-party workloads. “We went for 64-core AMD EPYC CPUs with 128 threads, to provide 200-plus virtual CPUs per server,” said Rainer Meier, System Architect, Telco Cloud Infrastructure. This enabled Swisscom to support more diverse workloads without expanding its footprint, driving greater efficiency, and enabling faster service delivery.

CHT ITG optimized its compute footprint by delivering more performance with fewer servers, especially for AI and virtual cloud environments. It was able to do this by deploying AMD EPYC™ processors with high core counts. “VMware benefits greatly from a high core count. It reduces the number of servers required and that means saving space and significantly reducing power consumption,” said Chung-Shuo Lin, Advisor, CHT ITG. This enabled the company to create a leaner, more energy-efficient infrastructure and one that was more prepared to handle future AI training and inference demands.

With high core counts and I/O, AMD EPYC CPUs empower telcos to deliver real-time responsiveness across core, edge, and RAN environments while enabling denser deployments and faster service delivery; essentially, enabling telcos to do more with less.

Powering Sustainability Goals

Environmental sustainability is a key area of focus for telcos. With rising energy costs, regulatory pressures, and enterprise commitments, leaders need infrastructures that align with long-term sustainability goals without sacrificing capability. Forward-looking leaders are already proving what is possible.

For example, Swisscom reduced energy consumption by 24% across 3,000 compute nodes, cutting per-vCPU power use by more than half. Similarly, CHT ITG reduced its rack space needs by 1.6x while increasing compute capacity thereby helping to achieve its carbon reduction goals while freeing up valuable space in its data center.

“The high performance of AMD EPYC processors allows the company to use fewer servers to provide higher computing capability, significantly optimizing hardware costs,” said Chung-Shuo Lin, Advisor, CHT ITG.

AMD EPYC CPUs are designed for energy efficiency while maintaining leadership performance. These benefits allow telcos to scale their server installations intelligently while supporting AI, cloud-native services, and 5G requirements. Deploying future-ready EPYC CPU-powered servers can help lower the total cost of ownership, compared to legacy installations, while helping ensure company infrastructure supports both traditional and emerging workloads into the foreseeable future.

Telco Infrastructure Agility Redefined

As telcos embrace network disaggregation and cloud-native principles, infrastructure agility is paramount. Leaders must manage diverse workloads ranging from traditional VNFs to containerized 5G cores and AI services across increasingly hybrid infrastructures. Business –ready systems today must be robust enough to support multiple, varied workloads without slowing down the infrastructure or making it unresponsive.

Operators are proving the business impact of getting agility right. Swisscom and CHT ITG each have streamlined operations by consolidating more workloads onto fewer, more capable systems—maximizing resource utilization, reducing operational overhead, and creating headroom for innovation. This approach has supported a wide mix of AI training, smart IoT, cloud services, and virtualized network cores while maintaining strong performance. “The hardware has way more performance now,” said Josua Hiller, Product Manager, Mobile Data Services, Swisscom, “so we can put many more workloads in one compute node.”

These implementations, from AMD processors, show how agile, well-managed infrastructures can sustain productivity and accelerate innovation. As telcos advance toward AI-driven service automation and intelligent networks, infrastructure agility will be a key competitive differentiator with faster service rollouts, reduced operational complexity, and long-term scalability.

Empowering the Future of Telecommunications Infrastructure

As telecommunications companies move from connectivity providers to digital service enablers, one thing is clear: AI is the next frontier. From network automation and anomaly detection to customer experience optimization and edge inferencing, AI is poised to transform telco operations and services.

Capturing that opportunity requires a flexible, AI-ready infrastructure that delivers performance where it matters most. In the data center, solutions built on AMD EPYC CPUs provide the compute headroom needed to deploy AI models, power 5G cores, and manage high-density workloads at scale, helping telcos reduce latency and optimize resources. Client systems featuring AMD Ryzen™ AI PRO processors enable protected, intelligent PCs that support enterprise AI applications, real-time collaboration, and next-gen productivity for a distributed workforce. At the edge, AMD technology-powered platforms bring inference, intelligent caching, and low-latency processing closer to customers, enhancing responsiveness, and enabling new services.

Together, these technologies give telco IT and business leaders the tools they need to help modernize infrastructures, streamline operations, and unlock new revenue opportunities while keeping pace with the AI revolution.

Discover how leading telco providers are reshaping their infrastructures to deliver higher performance, greater efficiency, and scalable growth, and explore more real-world success stories that can inspire your own transformation. 


Copyright © 2025 Advanced Micro Devices, Inc. Portions of this blog consist of AI-generated content. All rights reserved. AMD, the AMD Arrow logo, EPYC, AMD Instinct, Ryzen, and combinations thereof are trademarks of Advanced Micro Devices, Inc. Other product names contained herein are for identification purposes only and may be trademarks of their respective owners. Certain AMD technologies may require third-party enablement or activation. Supported features may vary by operating system. Please confirm with the system manufacturer for specific features. No technology or product can be completely secure.

Share:

Related Blogs