Samsung and AMD enter their long-term partnership in a more decisive stage and bring artificial intelligence further into commercial telecommunications. The revival energy between Samsung Electronics and Advanced Micro Devices is not just a technical upgrade. It represents an industry trend of wider infrastructure software-definition and AI-native that telecom operators need to scale, not merely experiment on in controlled laboratories. The shift of joint verification to actual rollouts is an important step, particularly at a time when operators are stressed to modernize at a cost and complexity-effective rate.
Samsung has also expanded its presence in North America in recent months with a large deployment with Videotron. The Samsung 5G Non-Standalone and 4G LTE Core gateway solutions run on CPU processors of AMD EPYC 9005 Series will be rolled out under the project. What is notable to industry observers is that such a deployment reflects the way high-end server grade CPUs have become the key to telecom core performance. To handle the workload, operators are shifting to high-performance processors in preference to each functional area needing a dedicated hardware device, such as traffic management, subscriber data processing, and artificial intelligence-driven analytics.
This strategy is consistent with a greater change in telecom architecture. Proprietary hardware was very important in traditional networks, and upgrades were tedious and expensive. In comparison, cloud-native cores and virtualized radio access networks enable functions to be implemented as software on standardized hardware. The EPYC processors by AMD are optimized to address the requirements of the intensive compute, thus fitting perfectly well in the context of the modern telecommunications infrastructure that place low latency, scalability, and energy efficiency as paramount features. Practically, it has the effect that operators can add capacity, add new services, and respond to new use-cases without re-creating their full hardware stack.
The collaboration is also becoming visible at the industry event like Mobile World Congress 2026 where Samsung is showing progress in AI-driven vRAN technology based on AMD processors. The two companies have multi-cell tested at Samsung in its research and development center after passing through previous validation milestones. The importance of these tests is that they would create conditions that would replicate the deployment conditions as opposed to remote laboratory testing environments. This is an indication that AI-enhanced vRAN systems are now prepared to be used in a larger scale since the stable, commercial-grade performance in multi-cell scenarios can be achieved.
Among the most striking details of this partnership, we can mention the fact that Samsung can provide AI-enhanced vRAN performance with the help of a completely virtualized software stack running on the latest AMD CPUs without any other hardware accelerators. Accelerators have been in demand in the telecom industry to match the exacting performance standards. Breaking that dependence may make network design easier and may also lower costs. It also opens more options to operators on how to scale and develop their infrastructure.
The achievement of Samsung with AMD highlights what can be achieved when AMD AI-native, open and virtualised architectures combine with innovative advances in compute, said Keunchul Hwang, Executive Vice President and Head of Technology Strategy Group, Networks Business, Samsung Electronics. We are moving forward to make AI-native networks be as scalable as possible to commercial- grade performance and be much more infrastructure optionable to make sure that their networks can be updated with the latest technologies and applications.
On the part of AMD, the partnership underscores the role that data center grade processors are playing in the next-generation connectivity. Telecommunications networks are not independent systems anymore. They become more like distributed cloud experiences, receiving large data streams of smartphones, IoT devices, enterprise applications, and edge applications. The computer intensity of such tasks demands processors that are able to support high throughput with low energy consumption.
Derek Dicker, corporate vice president, Enterprise Business Group said, 7 AMD EPYC CPUs are designed to support the high compute needs of contemporary telecommunications infrastructure. Our recent multi-cell vRAN performance with Samsung shows how our new generation EPYC processors provide the performance, efficiency and scalability that network operators and enterprises require to construct next-generation networks prepared to handle AI, automation and future developments.
In addition to core networks and vRAN, the collaboration is spread out to enterprise-oriented AI solutions. The Network in a Server by Samsung, which can be defined as a fully virtualized edge-AI solution based on AMD CPUs, is supposed to enable the operator to simplify the process of incorporating artificial intelligence into the network. The growing need of businesses to have high-speed processing of different applications like video analytics, industrial automation, and smart city deployments has made edge computing more relevant. The ability to combine both the network and AI in the server-based design allows the operators to minimize the complexity of the operations and provides new sources of revenue.
The example of AI-on-RAN use cases confirmed in the actual world setting in a large Japanese operator underlines the feasibility of the initiative. Some of the applications of these applications are video analysis, sensor and radar-based detection applications with integrated sensing and communication technology, and improved connectivity in next-generation devices. It is significant that it focuses on real-life validation. In telecom, the proof of concept is not all. Prolonged operation with live traffic is what defines the ability of a solution to scale.
The wider approach by Samsung is one based on creating an open ecosystem of partners in chipsets and not engaging operators on a one-stop vendor basis. When diversification of vendors is commonly regarded as a means of risk mitigation in an industry, the ability to offer a variety of hardware can build confidence in the operators. Simultaneously, AMD can enjoy the added integration into the portfolio of a global telecom vendor and expand its footprint on mobile infrastructure, instead of relying on the conventional data centers.
The Samsung Networks has been continuously increasing its portfolio over the last ten years, encompassing chipsets, radios, cloud-native cores, custom private networks, and AI-driven automation tools. The presence of the company in all the 5G stack places it in a strong position in the establishment of partnerships which cut across both hardware and software areas. Networks that operate under Samsung solutions are used by hundreds of millions of people worldwide, and introducing sophisticated AI functions to this footprint may affect how operators deal with such aspects of work as traffic optimization and predictive maintenance.
Nevertheless, the overall challenge of AI-native networks to the telecom industry is how they can bring technical potential to practical consumer and business advantages as fast as possible. Although cloud-native systems and virtual systems are flexible, they must also have well-trained operations departments and effective cybersecurity models. As the networks grow increasingly software-designed, they obtain the agility and the vulnerabilities of the cloud computing settings.
