Intel Launches Clearwater Forest “18A” Xeon+ Processors for 6G and Edge AI Applications

Intel has formally launched its next-generation Clearwater Forest “Xeon 6+”processors, utilizing cutting-edge 18A process technology designed to empower upcoming 6G technologies and Edge AI applications.

Revolutionizing 6G & Edge AI: Intel Clearwater Forest “Xeon 6+”CPUs

Press Release: As the industry gears up for 6G, it is increasingly evident that progress will stem from enhancing the robust computational foundation built during the 5G era. The focus will shift towards deploying scalable intelligence across existing infrastructures rather than complicating operations with new architectures. This vision is pivotal as we approach events like Mobile World Congress 2026.

Success in this next stage of network evolution will favor those who possess extensive industry experience and collaborative expertise. The challenge lies in simplifying network operations to enhance security and efficiency while ensuring AI inference aligns with real-world performance, power, and financial constraints.

Insights from Network Operators

Various themes frequently surface across the ecosystem as operators navigate the emerging landscape:

Integrating AI Seamlessly: Operators express a need for AI capabilities built directly into the network infrastructure instead of relying on disruptive accelerators or architectural revisions.
Prioritizing Efficiency: Operators aim to optimize power consumption and consolidate infrastructure, enabling new revenue streams while adapting to evolving user demands and usage patterns.
Fostering Trust Through Openness: There is a strong desire for secure, reliable, and open platforms that have a proven track record in commercial networks, ensuring a low-risk transition to 6G.

An Intel processor with a blue circuit design and intricate component layout is shown on a black background.

Intel is actively enhancing AI capabilities in the Radio Access Network (RAN) and Core with products like the Intel Xeon 6 with E-cores, Xeon 6 SoC, and the Intel Ethernet 800 and 600 series. The strategy is clear: to provide an open, secure computing foundation capable of running essential workloads—network functions, security protocols, enterprise applications, and AI inference—on a unified platform.

This robustness allows operators to modernize without the need for extensive overhauls, transforming their infrastructure into a catalyst for delivering faster services and improved economic efficiency. This approach not only benefits operators but also results in enhanced connectivity, personalized user experiences, and cost-effectiveness for consumers.

Choosing the Right Computing Architecture

While discussions often pit CPU against GPU, a binary viewpoint oversimplifies the complexities of network infrastructure evolution.

It is essential to pair AI workloads with the appropriate computational architectures rather than blindly applying a GPU-first approach. The Intel Xeon 6 with E-cores and Xeon 6 SoC can effectively increase network capacity and improve productivity while maintaining openness and control for operators.

Applying a GPU-centric mindset indiscriminately can drive up costs and operational complexity, create new silos, and necessitate architectural shifts that are often unwarranted. Therefore, in network operations, the real question shifts from “Can we run AI?” to “Can we execute AI efficiently without restructuring our existing systems, and what will it mean for our costs and power requirements in the future?”

In the RAN, the deployment of AI is about optimally matching workloads to the suitable computational resources. The Intel Xeon 6 SoC embeds AI acceleration directly into the Virtual RAN (vRAN) framework, leveraging Intel Advanced Matrix Extensions (AMX) and Intel vRAN Boost to manage the majority of inference workloads on the server without incurring the additional costs, energy consumption, or complexities of external AI hardware.

This innovative approach leads to tangible benefits: reduced total cost of ownership (TCO), better utilization of existing infrastructure, and AI that can be operationalized in real networks immediately, without needing significant architectural transformations. For operators focused on efficiency and cost, the Xeon 6 SoC enables AI deployment with predictable performance, streamlined operations, and scalable solutions across extensive networks.

Real-World Implementation of Intel Xeon 6 SoC

Here are several examples of how operators are currently leveraging the Intel Xeon 6 SoC:

Rakuten Mobile collaborates with Intel to utilize the integrated AI acceleration of the Intel Xeon 6 SoC, focusing on training, optimizing, and deploying sophisticated AI models that meet the rigorous demands of RAN workloads with ultra-low latency.
Vodafone has committed to employing Intel Xeon 6 SoCs for its large-scale Open RAN and vRAN upgrades across Europe, building on previous deployments in the UK where Intel Xeon facilitated its initial commercial Open RAN launches.

A gloved hand holds a silicon wafer with visible circuitry and components.

Additionally, here are instances of the Intel Xeon 6 with E-cores in action since its introduction over a year ago:

SK Telecom is implementing the Xeon 6 with E-cores and Intel Ethernet 800 Series in its mobile core production framework.
NTT DOCOMO has opted for the Xeon 6 with E-cores and Intel Ethernet E830 Network Adapter for its latest mobile core advancements.

Looking Towards the Future

As customer demands evolve toward platforms delivering reliable performance, scalability, and lower total costs, network equipment providers (NEPs) and service providers have begun to appreciate how Intel’s E-core architecture enhances efficiency, performance, and security in current Core infrastructure setups. The next phase in the Xeon 6 trajectory is now clear: Intel Xeon 6+.

A bar chart titled 'Ericsson Packet Core Using Intel Xeon 6+' explains the reduction in server count and energy consumption.

Engineered on Intel’s 18A technology, Xeon 6+ offers operators a scalable solution that dramatically minimizes energy consumption while facilitating enhanced intelligent network services. Its design focuses on improves core density and lowers energy expenditure, directly affecting total cost of ownership. As networks transition from 5G to cloud-centric applications, these processors are meticulously crafted to optimize efficiency, performance, and cost, fundamentally reshaping the landscape of data center economics in preparation for 6G. More insights regarding this next-generation technology will be shared as we approach MWC.

Source & Images