Posted on

Marvell and VMware Collaborate to Optimize the RAN

By Peter Carson, Senior Director Solutions Marketing, Marvell

and Tosin Olopade, Technical Product Line Manager, VMware

and Padma Sudarsan, Director of Engineering, RAN Architecture, VMware 

VMware, a pioneer in assisting communication service providers (CSPs) in transforming their networks, is partnering up with Marvell, a leading provider of data infrastructure semiconductor solutions to improve RAN performance and ROI. This collaboration provides solutions that enable CSPs to meet the demands of 5G’s increased capacity and use cases, optimizing the revenue and efficiency of each RAN site.

RAN sites worldwide are targeted for new technology deployment, where traditional, custom-made equipment is being replaced with servers adapted from data centers. This transformation to virtualized RAN and Open RAN, which replaces hardware with software, is driving the modernization of RAN sites worldwide. This allows CSPs to select servers and software based on their strategic goals, enabling them to offer unique services compared to their competitors. 

However, 5G RAN workloads, particularly Layer 1 (L1), are far more complex and latency-sensitive than the applications that general purpose CPUs have been designed to address. The load on even the most robust CPUs in the case of 5G RAN virtualization can be demanding. The rapid increase in 5G network speeds, reaching multi-gigabit-per-second, and the management of software-centric RAN Distributed Units (DUs) has resulted in rising energy consumption and cooling demands. This leads to increased costs, such as higher electricity bills, and may compromise CSPs’ plans to monetize their RAN investments. 

Marvell offers a solution to the challenges faced by the next generation of RAN with the OCTEON® 10 Fusion processor family for traditional base stations and virtualized Radio Access Network (vRAN) accelerator cards. With decades of experience in accelerating security, networking, and mobile innovations through data-centric processors, Marvell is bringing its expertise to the disaggregated RAN. The Marvell platform, offered as PCIe cards, takes on many demanding network-specific tasks currently handled by RAN server CPUs, providing open and L1-optimized hardware accelerators for common server architectures. 

VMware Telco Cloud Platform RAN™ is a cloud-native RAN solution that is designed specifically for running RAN functions. It provides the RAN modernization path, evolving from legacy RAN to vRAN to Open RAN. It transforms the RAN into a 5G multi-services hub “mini cloud”, enabling CSPs to monetize their RAN investments. VMware Telco Cloud Platform RAN is designed to meet the performance and latency requirements inherent to RAN workloads. 

The VMware Telco Cloud Platform is a common horizontal platform from Core to RAN for deploying 5G and traditional services in a multi-cloud environment. This common platform automatically self-tunes, depending on the workload that is being deployed through Telco Cloud Automation (TCA). 

The VMware Telco Cloud Platform accelerates and automates the provisioning and onboarding of vRAN functions across distributed RAN sites. 

By leveraging the platform, communication service providers can streamline operations from Day 0 to Day 2, reducing operational expenses and improving efficiency. The consistency of deployment across multiple sites further drives efficiency gains. The platform takes latency and bandwidth requirements into consideration when provisioning virtual RAN functions across thousands of platforms. 

VMware Telco Cloud Platform is comprised of the following:  

  • IaaS layer with VMware vSphere 
  • CaaS layer with VMware Tanzu Kubernetes Grid 
  • Automation layer with VMware Telco Cloud Automation performs CaaS automation and orchestration. 
  • Automated Workflow Engine for Hardware bootstrapping, OS Installation, BIOS configuration and Firmware upgrade with Bare Metal Automation. 
  • Assurance layer: VMware Telco Cloud Service Assurance handles FCAPS for the entire stack, along with the Aria Operations suite.  

VMware Telco Cloud Automation supports features such as: 

  • Photon OS, a real time Linux container host for DUs running on vSphere. 
  • Precision Time Protocol (PTP) accuracy 
  • Dynamic Infrastructure Policies programmatically optimizes your cloud resources by adjusting the underpinning infrastructure availability as you deploy network functions. 
  • Seamless, automated on-boarding and offloading full inline L1 PHY acceleration on Marvell® Smart NIC PCIe card. 

The Marvell OCTEON 10 Fusion 5G baseband processor—the industry’s first cutting edge 5nm technology—provides a range of accelerated functions. These include the complete handling of all L1 networking functions–physical and radio layer communications, including forward error correction, as well as Radio Resource Control (RRC) and Precision Time Protocol (PTP). 

Processing is performed in-line, meaning that all DU communications with the Radio Unit (RU) are delegated to the OCTEON 10 Fusion platform without adding any processing burden on the host. The result is faster network performance and greater CPU compute capacity while lowering overall server power consumption. 

The VMware Telco Cloud Platform will empower software vendors to run their RAN DU offerings on VMware’s virtualized and containerized infrastructure, taking advantage of Marvell’s acceleration features without resorting to low-level custom code. By expanding access to Marvell’s unique technologies, CSPs will have more options, leading to increased server efficiency, reduced energy consumption, adoption of edge computing, and RAN monetization.  

We are excited to collaborate with our partner ecosystem to bring this innovation to scale for communication service providers. 

Marvell disclaimer: This blog contains forward-looking statements within the meaning of the federal securities laws that involve risks and uncertainties. Forward-looking statements include, without limitation, any statement that may predict, forecast, indicate or imply future events or achievements. Actual events or results may differ materially from those contemplated in this blog. Forward-looking statements speak only as of the date they are made. Readers are cautioned not to put undue reliance on forward-looking statements, and no person assumes any obligation to update or revise any such forward-looking statements, whether as a result of new information, future events or otherwise. 

Comments are closed.