Intel unveils cutting-edge Xeon 6+ CPUs with 288 cores, targeting AI-ready networks

Intel Corp. has made a significant leap in the realm of artificial intelligence and data center technology with the introduction of its latest central processing unit (CPU), the Xeon 6+. This cutting-edge processor, unveiled at the Mobile World Congress (MWC) in Barcelona, is designed specifically for AI networks and other data center applications, marking a pivotal moment for the company as it seeks to redefine computing capabilities in the context of emerging technologies.

The Xeon 6+ CPU, codenamed "Clearwater Forest," is not just another addition to Intel's product line; it represents a sophisticated multi-chiplet design that integrates 12 compute tiles in an exceptionally dense core arrangement. This innovative architecture is tailored for modern networks and cloud infrastructures, particularly as the industry prepares for the transition to 6G technology. Intel has harnessed several advanced technologies, including high-bandwidth, on-chip fabric 2.5D Embedded Multi-Die Interconnect Bridge (EMIB) links and Foveros Direct 3D die stacking, to ensure that the Xeon 6+ can handle the demanding requirements of the new network standard.

Technical Specifications and Innovations

The Xeon 6+ CPU is built on Intel's most advanced 18A process node, which measures just 1.8 nanometers. This allows for the integration of 12 compute tiles alongside three active base tiles manufactured on the Intel 3 process (7nm) and two I/O tiles created on the Intel 7 process (10nm). Each compute tile can be further divided into six modules, each containing four Darkmont E-cores, culminating in a remarkable total of 288 cores per CPU. For dual-socket configurations, this number doubles to 576 cores, showcasing Intel's commitment to delivering high-performance computing.

Each core in the Xeon 6+ benefits from enhanced parallelism, improved vector throughput, and a larger instruction cache, which collectively contribute to a more efficient processing capability. The I/O tiles consist of two chiplets, each housing eight accelerators, alongside a robust set of interconnects: 48 PCIe Gen 5 lanes, 32 CXL 2.0 lanes, and 96 UPI 2.0 lanes. This architecture allows for seamless communication between multiple Xeon CPUs in large clusters, making it ideal for extensive data center operations.

Intel has also emphasized that the Xeon 6+ CPUs can support up to 12 memory channels and are compatible with DDR5-8000 modules, further enhancing their performance in data-intensive tasks. The new chips deliver double the core count compared to the previous Xeon 6700E CPUs, resulting in a 17% increase in instructions per clock per core, five times more last-level cache, and a 20% boost in memory speed.

Market Positioning and Future Prospects

Intel plans to launch the Xeon 6+ processors in the first half of the year, targeting network providers and data center operators as its primary market. The company envisions that cloud providers will be able to run dozens, if not hundreds, of virtual machines on a single Xeon 6+ CPU, significantly enhancing operational efficiency. In networking environments, these processors are specifically designed for applications such as Radio Access Network (RAN), 5G core, and edge workloads.

Kevork Kechichian, Intel’s executive vice president and general manager of the Data Center Group, highlighted the chip's capability to support real-time inference in virtualized RAN deployments. This feature allows data to be processed by AI models at the point of origin, minimizing the need to transfer data back and forth to cloud servers, thus streamlining operations and reducing latency.

Moreover, Intel is set to expand its collaboration with Swedish telecommunications giant Ericsson to co-develop and market "AI-native 6G solutions." Although specific details about this partnership remain limited, Kechichian described it as a step towards creating high-performance, energy-efficient computing architectures that cater to both AI applications and network functionalities. The concept of AI-native 6G aims to integrate intelligent and programmable networks with advanced computing and real-time sensing capabilities, paving the way for more responsive and efficient services.

Why it matters

The launch of the Intel Xeon 6+ CPU marks a significant milestone in the evolution of computing technology, particularly in the context of artificial intelligence and next-generation networking. For creators and technologists, this advancement suggests a future where AI applications can be deployed more efficiently and effectively, particularly in environments that require rapid data processing and real-time analytics.

The integration of advanced technologies such as 3D die stacking and multi-chiplet designs could lead to more powerful and versatile computing solutions, enabling developers to create innovative applications that leverage the full potential of AI. Additionally, the focus on AI-native 6G solutions indicates a growing trend towards the convergence of AI and networking, which could transform how data is processed and utilized across various industries.

As the demand for high-performance computing continues to rise, the Xeon 6+ CPU positions Intel as a key player in the market, offering tools that empower creators and technologists to push the boundaries of what is possible in AI and data-driven applications.