Intel accelerates 5G leadership

“Intel powers the world’s clouds, networks and enterprises, giving us unique insight on where to place compute and acceleration along the entire cloud-to-edge continuum and helping our customers scale to meet user demands,” said Sachin Katti, Intel senior vice president and general manager of the Network and Edge Group. “The advancements we’ve made in our 4th Gen Intel Xeon platforms to double vRAN performance1 while staying within the same power envelope, to nearly doubling the 5G core UPF throughput, and to speed the deployment of a wide range of network, security and enterprise edge services, makes Intel the platform for our customers to modernize and monetize their networks of the future, today.”

vRAN is Here and Nearly All Deployments Run on Intel

The need for high-performance, scalable, flexible and energy-efficient systems is driving the transformation of mobile networks from fixed function, hardware-based silicon and infrastructure to software-based, fully virtualized platforms running on general-purpose processors. Accelerating the virtualization of the RAN positions communications service providers (CoSPs) to meet future requirements while improving RAN energy efficiency and reducing their total cost of ownership (TCO).

With expansive industry support from Advantech, Capgemini, Canonical, Dell Technologies, Ericsson, Hewlett Packard Enterprise, Mavenir, Quanta Cloud Technology, Rakuten Mobile, Red Hat, SuperMicro, Telefonica, Verizon, VMware, Vodafone and Wind River, among others, Intel launched 4th Gen Intel Xeon Scalable processors with Intel vRAN Boost.

By fully integrating vRAN acceleration into the Intel Xeon system-on-chip (SoC) and eliminating the need for an external accelerator card, Intel is delivering 2x the capacity gains gen over gen within the same power envelope and up to an additional 20% power savings with integrated acceleration, which is above and beyond the 4th Gen Intel Xeon platform’s already outstanding performance-per-watt gain. With this combination of processing innovations and feature integration, Intel expects 4th Gen Xeon Scalable processors with Intel vRAN Boost will match or better the performance-per-watt of the best Layer 1 SoC accelerator cards in the market today, while delivering the benefits of software-defined, virtualized networks.

5G Core Networks Run on Software, Delivering Cloud-Native Agility

At the core of the network, Intel is leading the evolution to cloud-native, service-based architectures with open solutions for addressing challenges like performance, TCO, power efficiency, security and lack of visibility across the network stack. Intel’s hardware and software solutions will enable 5G core networks to work harder and smarter at achieving a balance between critical business and customer requirements for power efficiency, performance and latency.

To further assist network operators in modernizing their networks, reducing their total cost of ownership (TCO) at the 5G core, Intel demonstrated 4th Gen Intel Xeon Scalable processors now enable the industry’s first 1 Tbps of performance for the 5G UPF workload within a single dual-socket server, further validated by Samsung.

Additionally, the new Intel Infrastructure Power Manager for 5G Core reference software dynamically matches run-time server power consumption with data traffic without compromising key performance indicators such as throughput, latency and packet drop.

The software, in tests with Casa Systems, NEC and Nokia, significantly reduces time-to-market for ISVs and operators by simplifying access to key capabilities in Intel’s 3rd and 4th Gen Xeon Scalable processors, including power telemetry, granular power control states and low-latency frequency change. Operators can use the reference software to reduce network TCO and accelerate progress toward net zero emissions goals, realizing millions of dollars in potential savings and a significant amount of CO2 emissions offset.

Intel is Setting the Pace at the Edge

Massive growth at the network edge, largely in video services, will define much of this decade’s competitive landscape for service providers. The operators’ network edge facilities position them for a competitive advantage to deliver to this growth, yet it’s challenging to predict which specific video services will take off.

Alongside partners Broadpeak, China Mobile, Cloudsky, Thundersoft and ZTE, Intel showcased the Intel Converged Edge Media Platform, which delivers multiple video services from a shared, multitenant architecture and leverages cloud-native scalability to intelligently respond to shifting requirements. 

Video services – such as CDNs (content delivery networks), cloud gaming, mixed reality and 3D rendering – can be delivered in a single cloud-native environment supported by both CPU and GPU-accelerated applications. Operators no longer need to invest in dedicated resources for services that may not take off. Instead, they can build on a general-purpose architecture where services reside together and can leverage cloud-native scalability to automatically change or resize services to changing needs.

Delivering Customer Choice in Acceleration

Alongside the integrated network acceleration built into 4th Gen Intel Xeon Scalable processors, Intel is expanding its Agilex 7 FPGAs and eASIC N5X structure ASIC devices for cloud, communications and embedded applications. 

As cloud service providers (CSPs) begin transitioning from 200G to 400G networks in 2023, with CoSPs to follow suit in 2024, Intel Agilex 7 FPGA AGI 041 devices will enable next-generation 400G infrastructure acceleration solutions. AGI 041 devices deliver the right balance of capacity, power efficiency and performance for the 400G Infrastructure Processing Unit (IPU) and networking solutions.

Additionally, Intel provides the unique ability to further optimize cost and power across customers’ 400G infrastructure solutions through Intel eASIC structured ASICs. For networking workloads, N5X080 devices are capable of reducing core power by up to 60% versus an FPGA, while reducing prototyping time by 50% compared to a traditional ASIC.