Ericsson
AT&T and Ericsson boost Cloud RAN performance with AI-native software running on Intel Xeon 6 SoC
Overview:
AT&T and Ericsson have completed a milestone Cloud RAN test by successfully demonstrating Ericsson’s AI-native Link Adaptation [1.] on a Cloud RAN stack powered by Intel Xeon 6 SoC. The test showed how artificial intelligence (AI) can improve spectral efficiency and network responsiveness in real-world conditions. Conducted over AT&T’s licensed frequency bands, the experiment was the first to use portable Ericsson RAN software running on Intel’s new Xeon 6 system-on-chip (SoC) platform—an architecture designed for high-performance, cloud-native processing of RAN workloads. Engineered specifically for network and edge deployments, Intel Xeon 6 SoC delivers breakthrough AI RAN performance with built-in acceleration. Integrated Intel Advanced Vector Extensions (AVX) and Intel Advanced Matrix Extension (AMX) technologies eliminate the need for discrete accelerators while maximizing capacity, efficiency, and TCO optimization.
…………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………..
Note 1. AI-native Link Adaptation dynamically adjusts to changes in signal quality and interference, boosting RAN performance on purpose-built and cloud-based infrastructure alike.
Other Notes:
-
vRAN: A radio access network (RAN) in which the baseband processing functions run as software on general-purpose processors (mostly from Intel) instead of on dedicated hardware at the cell site. In vRAN, the functional split defines how baseband processing is divided between centralized processors and the radio unit at the site, and that split drives fronthaul bandwidth, latency, and cost.
- Cloud RAN: An evolution of vRAN where those same RAN functions are re-architected as cloud‑native microservices/containers with CI/CD (Continuous Integration and either Continuous Delivery or Continuous Deployment), automation, and orchestrators, optimized for elastic scaling across distributed cloud infrastructure.
- Ericsson Cloud RAN is a cloud native software solution that handles compute functionality in the RAN. It virtualizes RAN functions on Commercial Off The Shelf (COTS) hardware, decoupling software from hardware to enable more flexible, scalable, and efficient network deployments.
- According to Dell’Oro Group, Cloud RAN (often encompassing vRAN) accounted for approximately 5% to 10% of the total global Radio Access Network (RAN) market revenues in 2025. In early 2026, Dell’Oro revised Cloud RAN projections downward. While virtualization remains a “key pillar” for the long term, short-term adoption is being slowed by performance, power, and cost-parity challenges when compared to purpose-built hardware.
- The total RAN market stabilized in late 2025 after losing approximately 20% of its value between 2022 and 2024. Market concentration reached a 10-year high in 2025, with the top five vendors (Huawei, Ericsson, Nokia, ZTE, and Samsung) capturing 96% of the revenue.
………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………..
Image Credit: Ericsson
In this proof-of-concept setup, Ericsson’s disaggregated and containerized RAN software operated within AT&T’s target Cloud RAN configuration, built on open, commercial off-the-shelf hardware. The test advanced from basic call functionality to validation of feature-rich network behavior in a cloud computing environment. Ericsson’s AI-native Link Adaptation is a learning algorithm that continuously assesses channel state and interference to determine the optimal modulation and coding scheme for each transmission interval. By generating real-time predictions of link quality, the AI model dynamically adjusts data rates to maximize throughput and spectral efficiency.
Early results were promising. Throughput gains reached up to 20% compared with conventional rule-based link adaptation approaches, alongside measurable improvements in spectral efficiency. Ericsson and Intel also used the trial to benchmark various AI inference models, demonstrating performance scalability and energy efficiency on general-purpose compute nodes rather than proprietary hardware accelerators. This suggests a more pragmatic path for deploying AI workloads across distributed RAN architectures.
AI-native Link Adaptation dynamically adjusts to changes in signal quality and interference, boosting RAN performance on purpose-built and cloud-based infrastructure alike.
Ericsson Cloud RAN is a cloud native software solution that handles compute functionality in the RAN. It virtualizes RAN functions on Commercial Off The Shelf (COTS) hardware, decoupling software from hardware to enable more flexible, scalable, and efficient network deployments.
Engineered specifically for network and edge deployments, Intel Xeon 6 SoC delivers breakthrough AI RAN performance with built-in acceleration. Integrated Intel Advanced Vector Extensions (AVX) and Intel Advanced Matrix Extension (AMX) technologies eliminate the need for discrete accelerators while maximizing capacity, efficiency, and TCO optimization.
Beyond the immediate performance improvements, the trial illustrates how open RAN architectures can accelerate innovation. By decoupling RAN software from vendor-specific hardware, AT&T can integrate AI capabilities and update network functions more quickly, avoiding the constraints of lock-in. The portability demonstrated here—running production-grade Ericsson RAN software on Intel Xeon 6 silicon—marks an industry first.
For AT&T, the achievement represents more than a lab milestone. It provides a technical template for scaling AI-native RAN functions into its cloud infrastructure, pointing to a future where machine learning operates natively within radio environments to fine-tune performance in real time. As operators continue balancing cost, flexibility, and efficiency, AI-optimized Cloud RAN deployments could become the next competitive frontier in 5G—and eventually, 6G—network evolution.
………………………………………………………………………………………………………………………………………………………………………………………………………………………..
Quotes:
Rob Soni, Vice President, RAN Technology at AT&T, says: “AT&T is leading the charge toward an open, intelligent, and scalable network future by advancing Open RAN and Cloud RAN with AI-native capabilities at their core. This demo highlights how AI capabilities, powered by our next-generation Cloud RAN platform, can be deployed seamlessly to drive innovation and deliver superior customer experiences.”
Mårten Lerner, Head of Networks Strategy and Product Management, Business Area Networks at Ericsson, says: “Together with AT&T and Intel, Ericsson is demonstrating how our domain expertise combined with AI-native RAN software can drive transformative advancements in both Cloud RAN and purpose-built deployments. Our industry-leading AI-native Link Adaptation serves as the first proof point on this journey. With a hardware-agnostic RAN software stack, Ericsson is committed to offering maximum flexibility and enabling all our customers to benefit from future innovations – regardless of their chosen underlying hardware. This milestone underscores Ericsson’s commitment to helping operators advance their networks by deploying AI functionality across the RAN stack.”
Cristina Rodriguez, VP and GM, Network and Edge at Intel, says: “This successful collaboration with AT&T and Ericsson showcases the power of Intel Xeon 6 SoC to enable and accelerate AI workloads in Cloud RAN environments. Xeon 6 SoC is architected to handle the demanding compute requirements of AI-native network functions, delivering the performance and efficiency operators need to unlock the full potential of intelligent networks. By providing a flexible, standards-based platform, Intel Xeon 6 enables service providers like AT&T to deploy innovative AI capabilities while maintaining the openness and choice that drive industry innovation.”
………………………………………………………………………………………………………………………………………………………………………………………………………………………….
AI-Native Link Adaptation vs. Traditional Methods:
Traditional link adaptation in RAN relies on deterministic, rule-based algorithms that select the Modulation and Coding Scheme (MCS) from predefined lookup tables. These methods primarily use instantaneous Channel Quality Indicator (CQI) reports or estimated Signal-to-Interference-plus-Noise Ratio (SINR) thresholds, often adjusted via Outer Loop Link Adaptation (OLLA) based on ACK/NACK feedback from the UE. This reactive approach applies conservative margins to account for channel estimation errors, prediction lag, and varying interference, which can lead to suboptimal throughput—either underutilizing the link with low MCS or triggering excess HARQ retransmissions with overly aggressive selections.
AI-native Link Adaptation shifts to a predictive, model-driven paradigm using machine learning (typically lightweight neural networks or time-series models) trained on historical channel data. Rather than static thresholds, the AI processes sequences of CQI, beam metrics, mobility patterns, and interference traces to forecast the probable channel state for the next transmission time interval (TTI). This enables precise MCS selection that hugs the Shannon capacity limit more closely, minimizing BLER while maximizing spectral efficiency in dynamic scenarios like high-mobility NLOS or bursty interference.
Key differences include:
| Aspect | Traditional (Rule-Based) | AI-Native (ML-Based) |
|---|---|---|
| Decision Mechanism | Lookup tables, SINR thresholds, OLLA offsets | Real-time inference from ML models |
| Channel Handling | Reactive (past CQI/SINR) | Predictive (time-series forecasting) |
| Adaptation Speed | Step-wise, with feedback lag | Continuous, sub-TTI granularity |
| Performance Gains | Baseline (0% reference) | Up to 20% throughput, 10% spectral efficiency |
| Compute Needs | Low (fixed arithmetic) | Moderate (edge inference on COTS like Xeon 6) |
| Limitations | Struggles with non-stationary channels | Requires training data, retraining overhead |
Analysis: Rakuten Mobile and Intel partnership to embed AI directly into vRAN
RAN silicon rethink – from purpose built products & ASICs to general purpose processors or GPUs for vRAN & AI RAN
vRAN market disappoints – just like OpenRAN and mobile 5G
Nokia and Eolo deploy 5G SA mmWave “Cloud RAN” network
Ericsson and Google Cloud expand partnership with Cloud RAN solution
Ericsson and O2 Telefónica demo Europe’s 1st Cloud RAN 5G mmWave FWA use case
Cloud RAN with Google Distributed Cloud Edge; Strategy: host network functions of other vendors on Google Cloud
vRAN market disappoints – just like OpenRAN and mobile 5G
Ericsson and Intel collaborate to accelerate AI-Native 6G; other AI-Native 6G advancements at MWC 2026
Ericsson and Intel at MWC 2026:
Building on milestones in Cloud RAN, 5G Core, and open network innovation, Ericsson and Intel are showcasing joint technology advancements at the Mobile World Congress (MWC) 2026 in Barcelona this week. Demonstrations can be experienced at the Ericsson Pavilion (Hall 2), Intel Booth (Hall 3, Stand 3E31), and across partner event spaces, highlighting the companies’ shared progress in enabling the next era of AI-driven networks.
The two companies are strengthening their long-standing technology partnership to accelerate ecosystem readiness for AI-native 6G networks and use cases. The expanded collaboration spans next-generation mobile connectivity, cloud infrastructure, and compute acceleration — with a focus on AI-driven RAN and packet core evolution, platform-level security, and scalable cloud-native architectures designed to shorten time-to-market for advanced network solutions.
“6G is not merely an iteration of mobile technology; it will serve as the foundational infrastructure distributing AI across devices, the edge, and the cloud,” said Börje Ekholm, President and CEO of Ericsson. “With our deep history in network innovation and global-scale operator deployments, Ericsson is uniquely positioned to drive practical 6G integration from research to commercialization.”
Lip-Bu Tan, CEO of Intel, added: “Intel’s vision is to lead the industry in unifying RAN, Core, and edge AI to enable seamless deployment of AI-native 6G environments. Together with Ericsson, we are proving that next-generation connectivity can be open, energy-efficient, secure, and intelligent. With future Ericsson Silicon built on Intel’s most advanced process technologies, coupled with Intel Xeon-powered AI-RAN ready Cloud RAN and collaborative multi-year research efforts, we are delivering the performance, efficiency, and supply assurance demanded by leading operators worldwide.”
As 6G transitions from research to commercialization, the industry must align around a mature, standards-based ecosystem. The Ericsson–Intel collaboration aims to accelerate development of high-performance, energy-efficient compute architectures optimized for both AI for Networks and Networks for AI.
AI-native 6G will fuse intelligent, programmable network functions with distributed compute and real-time sensing, bringing processing power closer to the network edge and enabling ultra-responsive, adaptive services. This convergence will enhance network efficiency, agility, and service intelligence across future deployments.
About Ericsson:
Ericsson‘s high-performing networks provide connectivity for billions of people every day. For 150 years, we’ve been pioneers in creating technology for communication. We offer mobile communication and connectivity solutions for service providers and enterprises. Together with our customers and partners, we make the digital world of tomorrow a reality.
About Intel:
Intel is an industry leader, creating world-changing technology that enables global progress and enriches lives. Inspired by Moore’s Law, we continuously work to advance the design and manufacturing of semiconductors to help address our customers’ greatest challenges. By embedding intelligence in the cloud, network, edge and every kind of computing device, we unleash the potential of data to transform business and society for the better.
…………………………………………………………………………………………………………………………………………………………
Related AI-Native 6G Announcements at MWC 2026:
In addition to the Ericsson-Intel collaboration, several vendors and operators announced AI-native 6G advancements or related demos at MWC Barcelona 2026. These initiatives emphasize AI-RAN integration, software-defined architectures, and ecosystem partnerships to bridge 5G-A to 6G.
NVIDIA Multi-Partner Commitment: NVIDIA rallied operators and vendors including Booz Allen, BT Group, Cisco, Deutsche Telekom, Ericsson, Nokia, SK Telecom, SoftBank, and T-Mobile to build open, secure AI-native 6G platforms. The focus is on software-defined wireless with AI embedded in RAN, edge, and core for integrated sensing, communications, and interoperability.
Nokia AI-RAN: Nokia highlighted new partnerships with Dell, Quanta, Red Hat, SuperMicro, NVIDIA, and operators like T-Mobile, Indosat Ooredoo Hutchison, BT, Elisa, NTT DOCOMO, and Vodafone for AI-RAN trials paving the way to cognitive 6G networks. Live demos at Nokia’s Hall 3 Booth 3B20 included Southeast Asia’s first AI-RAN Layer 3 5G call on shared GPU infrastructure and vision AI for immersive services.
T-Mobile & Deutsche Telekom Hub: T-Mobile US and (major shareholder) Deutsche Telekom launched a joint 6G Innovation Hub targeting AI-native autonomous networks, secure sensing/positioning, and connectivity-compute convergence for Physical AI. It builds on agentic AI proofs like network-integrated translation, emphasizing “kinetic tokens” for real-time physical world control.
ZTE GigaMIMO 6G Prototype: ZTE unveiled the world’s first 6G prototype with 2000+ U6G-band antenna elements (GigaMIMO), powered by AI algorithms for 10x capacity over 5G-A, 30% spectral efficiency gains, and AI-driven immersive services. Booth 3F30 demos integrate AI across connectivity, computing, and devices for “AI serves AI” networks.
Qualcomm Agentic AI RAN: Qualcomm announced AI-native RAN management services in its Dragonwing suite for autonomous 6G-grade networks, plus new Open RAN AI features for performance optimization. CEO Cristiano Amon’s keynote focused on “Architecting 6G for the AI Era,” with device-to-data-center transformations.
Huawei U6GHz for 6G Path:
Huawei released all-scenario U6GHz products (macro/micro sites, microwave) with AI-centric solutions for 5G-A capacity (100 Gbps downlink) and low-latency AI apps, enabling smooth 6G evolution. Emphasizes hyper-resolution MU-MIMO and multi-band coordination for indoor/outdoor AI experiences.
Summary Chart:
| Vendor/Operator | Key Focus | Partners/Demos | Booth/Location |
|---|---|---|---|
| NVIDIA | Open AI-native platforms | Multiple operators/vendors | MWC general |
| Nokia | AI-RAN trials & cognitive networks | NVIDIA, T-Mobile, IOH et al. | Hall 3, 3B20 |
| T-Mobile/DT | Physical AI hub | Joint R&D | Announced pre-MWC |
| ZTE | GigaMIMO 6G prototype | China Mobile, Qualcomm | Hall 3, 3F30 |
| Qualcomm | Agentic RAN automation | Open RAN ecosystem | Keynote & demos |
| Huawei | U6GHz AI-centric evolution | Carrier-focused | MWC showcase |
…………………………………………………………………………………………………………………………………………………………………………………….
References:
NVIDIA and global telecom leaders to build 6G on open and secure AI-native platforms + Linux Foundation launches OCUDU
Comparing AI Native mode in 6G (IMT 2030) vs AI Overlay/Add-On status in 5G (IMT 2020)
SKT 6G ATHENA White Paper: a mid-to-long term network evolution strategy for the AI era
Dell’Oro: RAN Market Stabilized in 2025 with 1% CAG forecast over next 5 years; Opinion on AI RAN, 5G Advanced, 6G RAN/Core risks
Nokia and Rohde & Schwarz collaborate on AI-powered 6G receiver years before IMT 2030 RIT submissions to ITU-R WP5D
SK Telecom, DOCOMO, NTT and Nokia develop 6G AI-native air interface
Market research firms Omdia and Dell’Oro: impact of 6G and AI investments on telcos
Ericsson goes with custom silicon (rather than Nvidia GPUs) for AI RAN
Dell’Oro: Analysis of the Nokia-NVIDIA-partnership on AI RAN
RAN silicon rethink – from purpose built products & ASICs to general purpose processors or GPUs for vRAN & AI RAN
Nokia to showcase agentic AI network slicing; Ericsson partners with Ookla to measure 5G network slicing performance
Executive Summary:
Today, Nokia announced a strategic collaboration with Amazon (AWS), Du, and Orange to debut an industry-first agentic AI-driven network slicing [1.] capability on a 5G SA core network. Du and Orange will deploy this new technology which uses Nokia’s 5G AirScale base stations, MantaRay SMO and Agentic AI modules in tandem with Amazon’s Bedrock Artificial Intelligence platform. Autonomous AI agents are used to ingest and process real-time telemetry—including geospatial data, event triggers, and traffic patterns—the framework enables adaptive network slicing. This architecture allows communications service providers (CSPs) to dynamically orchestrate resources in response to fluctuating demand, such as prioritizing mission-critical throughput for first responders during emergency incidents.
Note 1. There are no ITU standards for network slicing or the 5G SA Core network required to implement that capability. 3GPP specifications define end-to-end network slicing architecture, covering slice management (TS 28.552, TS 28.554), service requirements, and security (NSSAA – Network Slice Specific Authentication and Authorization). The NSA and CISA have released specific, recognized guidance on designing, deploying, and maintaining secure 5G standalone (SA) network slices. ETSI publishes and adopts 3GPP technical specifications (specifically the 28-series) as European standards for network slicing management, including 5G RAN, core network, and NFV-MANO architecture. ETSI, as a 3GPP partner, ensures these specifications cover the lifecycle of network slices.
…………………………………………………………………………………………………………………………………………………………………………………………..
- Data Ingestion & Inference: Agentic AI modules, hosted on Amazon Bedrock, ingest real-world contextual data (e.g., emergency alerts, traffic sensors, weather) alongside live network KPIs.
- Intent-Based Policy Generation: The AI agents analyze this telemetry to determine the optimal network configuration required to meet specific Service Level Agreements (SLAs) or emergency “intents'”
- NEF & SMO Integration: These high-level intents are translated into actionable policies and pushed to Nokia’s MantaRay SMO (Service Management and Orchestration).
- Dynamic RAN/Core Adjustment: The Network Exposure Function (NEF) acts as the secure gateway, allowing the AI agents to interface with the 5G Core. It exposes network capabilities so the agents can dynamically adjust RAN policies and resource allocation across the 5G AirScale base stations.
- Autonomous Feedback Loop: The system operates in an autonomous mode where agents continuously monitor the results of their adjustments, performing forensic analysis to refine slicing parameters in real-time.
Nokia will host live technical demonstrations of this AI network slicing capability at its 2026 Mobile World Congress (MWC) Barcelona exhibit.
Quotes:
“This innovation marks a major milestone in the evolution of AI-native networks,” said Pallavi Mahajan, Chief Technology and AI Officer at Nokia. “By combining Nokia’s advanced network slicing capabilities with agentic AI, we are enabling operators to deliver premium, intent-based services that adapt dynamically to real-world conditions. Nokia is advancing connectivity by unlocking new value streams for telecommunication providers and supporting next-generation applications and differentiated services for enterprises, industries and consumers.”
Amir Rao, Global Director, GTM & Telco Solutions at AWS added: “Network slicing has long promised to unlock new revenue streams for operators, but manual configuration and static policies have prevented end customers from accessing on-demand provisioning. By integrating agentic AI capabilities through Amazon Bedrock with Nokia’s application, operators can now deliver intelligent, context-aware network slicing that responds dynamically to real-world conditions from traffic surges to emergency situations. This transforms network slicing from a technical capability into a true business enabler, allowing operators to monetize their 5G investments through differentiated, premium services that adapt automatically to customer needs. Agentic Network Slicing is the beginning of an era that will enable telecommunications providers to enable real-time intent-based service provisioning for end customers.”
………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………..
Competitive Network Slicing Solution:
Rival wireless equipment vendor Ericsson yesterday gave a preview of a network slicing related offering which it will be demonstrating at the 2026 MWC. Together with Ookla it has developed a specialized test version of its Speedtest app designed to measure and validate 5G network slicing performance. The tool enables the Speedtest app to identify and test specific network slices, which apparently demonstrates how Service Level Agreements (SLAs) for differentiated services can be verified in real-time by consumers and service providers.
Ericsson reported in its latest Mobility report that there were 65 commercial network slicing services worldwide providing so-called “differentiated connectivity” offerings. That’s out of a total of 118 network slicing cases discovered by Ericsson’s researchers. Yet in the UK, none of the three mobile network operators have launched a commercial 5G network slicing capability yet. According to Ofcom’s latest Connected Nations report, 5G SA is available across 83% of outside areas in the country and 5G SA accounts for nearly one-third of 5G traffic. However, 4G accounts for 72% of total monthly data traffic.
“Network slicing is no longer a future concept; it is a commercial reality. However, you cannot manage what you cannot measure,” said Tibor Rathonyi, Senior Advisor at Ookla. “Our work with Ericsson is a pivotal first step in providing the transparency needed to prove the value of these premium 5G services to both consumers and enterprises.”
Philipp Bichsel, Executive Vice President Mobile Network & Services at Swisscom, said: “Swisscom has retained the title as the country’s best-performing mobile network over many years by truly prioritizing the delivery of the best possible customer experience. This has meant embarking on a journey to fully exploit automation to enhance reliability and efficiency without compromising the service quality our customers expect. As we advance towards self-learning, autonomous networks, enabling Swisscom to build smarter and more adaptive network operations, we are leveraging the SMO framework as the foundation for this evolution. Within this framework, partner solutions such as Ericsson’s Intelligent Automation Platform and its ecosystem of rApps play an important role in helping us explore the potential of AI driven automation.”
…………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………..
References:
https://www.telecoms.com/5g-6g/nokia-and-aws-show-off-agentic-ai-powered-5g-advanced-network-slicing
https://www.telecoms.com/5g-6g/ericsson-and-ookla-launch-network-slicing-measurement-tool
https://www.lightreading.com/5g/eurobites-network-slicing-enjoying-a-moment-finds-ericsson-report
https://www.ericsson.com/en/reports-and-papers/mobility-report/reports/november-2025
https://www.lightreading.com/5g/5g-network-slicing-not-ready-for-prime-time-in-uk
https://www.awardsolutions.com/portal/resources/network-slicing
ABI Research: 5G network slicing market to hit $67.52 billion in 2030 with Asia Pacific in the lead
5G network slicing progress report with a look ahead to 2025
FCC Draft Net Neutrality Order reclassifies broadband access; leaves 5G network slicing unresolved
Telstra achieves 340 Mbps uplink over 5G SA; Deploys dynamic network slicing from Ericsson
ABI Research: 5G Network Slicing Market Slows; T-Mobile says “it’s time to unleash Network Slicing”
Ericsson, Intel and Microsoft demo 5G network slicing on a Windows laptop in Sweden
Ericsson and Nokia demonstrate 5G Network Slicing on Google Pixel 6 Pro phones running Android 13 mobile OS
Nokia and Safaricom complete Africa’s first Fixed Wireless Access (FWA) 5G network slicing trial
Is 5G network slicing dead before arrival? Replaced by private 5G?
5G Network Slicing Tutorial + Ericsson releases 5G RAN slicing software
Ericsson goes with custom silicon (rather than Nvidia GPUs) for AI RAN
Ahead of MWC Barcelona 2026, Ericsson unveiled its initial suite of AI-RAN products at a pre-event briefing in London, emphasizing a strategy anchored in proprietary, purpose-built silicon to enhance radio access network (RAN) performance. While the wireless industry is finally moving to virtualized/cloud RAN utilizing general-purpose processors from Intel, Ericsson is defending its continued investment in custom silicon for specialized, high-performance tasks.
Concurrently, the company is demonstrating a strong push toward software-defined flexibility, ensuring its proprietary RAN algorithms and AI-native software are portable across diverse, open silicon platforms. Ericsson was exploring the use of Nvidia’s Arm-based Grace CPU, rather than the Hopper-branded GPU, but has opted for custom silicon (ASICs) instead.
Ericsson’s RAN portfolio currently diverges into two primary architectures. The majority of its footprint relies on ASICs—developed through internal design and external partnerships with Intel. The alternative is Cloud RAN, which pairs Ericsson’s software stack with Intel Xeon processors. Despite the industry’s promise that virtualization would decouple hardware from software, Intel remains Ericsson’s sole silicon partner for production-grade deployments.
This hardware lock-in was underscored during Ericsson’s recent London event, where documentation confirmed “commercial support” exclusively for Intel, while AMD, Arm, and NVIDIA remain relegated to “prototype support.” Despite years of industry rhetoric regarding silicon diversity in the vRAN ecosystem, tangible progress remains stalled. Furthermore, the integration of AI into RAN software introduces new layers of complexity that may further entrench hardware dependencies.
Industry observers remain skeptical of Ericsson’s ambition for a “unified software stack” across heterogeneous hardware platforms. While hardware-software disaggregation is achievable in the higher layers (L2/L3), Layer 1 (L1)—the most compute-intensive portion of the stack—remains heavily optimized for specific silicon. Ericsson’s initial strategy relied on the portability of L1 code across x86 architectures (via AMD) or the adoption of Arm’s SVE2 (Scalable Vector Extension) to match Intel’s AVX-512 capabilities. However, achieving high-performance parity across these platforms without significant refactoring remains a significant engineering hurdle.
A critical bottleneck in PHY Layer (L1) processing is Forward Error Correction (FEC), which traditionally necessitates dedicated hardware acceleration. Ericsson initially addressed this using a lookaside acceleration model, offloading FEC tasks to discrete PCIe-based Intel accelerators. In recent iterations, Intel has moved toward a more integrated System-on-Chip (SoC) approach, embedding the accelerator directly onto the CPU die (e.g., vRAN Boost).
The primary challenge for Ericsson lies in achieving silicon parity across the AMD and NVIDIA ecosystems. While AMD’s FPGA-based accelerators have faced scrutiny regarding power efficiency, NVIDIA’s GPU-based offloading was previously viewed as cost-prohibitive for standard FEC. However, the rise of AI-RAN has recalibrated these economic models, as telcos explore the dual-use potential of GPUs for both RAN and AI workloads. Emerging platforms, such as Google’s Tensor Processing Units (TPUs), have also been identified by Ericsson leadership as viable long-term options.
Despite ambitions for a unified “single software track,” Ericsson’s technical roadmap suggests a more nuanced reality. While L2 and higher layers aim for a universal codebase across hardware platforms, L1 necessitates concurrent feature development and platform-specific tailoring. As CTO Erik Ekudden noted, maximizing the efficiency of advanced accelerators requires a degree of software customization that challenges the ideal of total hardware-software disaggregation.
Ericsson executives are keen to avoid what Executive VP Per Narvinger describes as a “native implementation,” which would create silicon vendor lock-in. To prevent that the company is prioritizing Hardware Abstraction Layers (HALs). Key initiatives include the adoption of the BBDev (Baseband Device) interface to decouple RAN software from underlying silicon. Furthermore, potential integration with NVIDIA’s CUDA platform is being evaluated to provide the necessary abstraction for heterogeneous compute environments, though this remains contingent on broader industry standardization.
Ericsson’s AI strategy mirrors this modular approach. By leveraging AI as a functional abstraction layer, the company aims to simplify software portability across diverse platforms while maintaining AI control loops for real-time network management. Unlike competitors tethered to high-TDP GPUs, Ericsson maintains that AI-RAN is commercially viable on general-purpose and purpose-built silicon. Recent London showcases highlighted AI-driven gains in spectral efficiency, channel estimation, and beamforming without external acceleration. A production-ready AI-native link adaptation model recently delivered a 10% spectral efficiency improvement in field tests and is now integrated into the latest baseband portfolio.
As for radios—a domain less susceptible to full virtualization—Ericsson is embedding Neural Network Accelerators (NNA) directly into its radio-unit ASICs. These programmable matrix cores are optimized for Massive MIMO inference, enabling sub-millisecond beamforming and channel estimation while adhering to strict site power envelopes. New AI‑ready radios, feature Ericsson custom silicon with neural network accelerators. They are said to boost on‑site AI inference capabilities in Massive MIMO radios, enabling real‑time optimization and full stack, fully distributed AI.
………………………………………………………………………………………………………………………………………………………………………………………………………………
References:
https://www.lightreading.com/5g/ericsson-does-ai-ran-minus-nvidia-in-push-for-5g-silicon-freedom
RAN silicon rethink – from purpose built products & ASICs to general purpose processors or GPUs for vRAN & AI RAN
Analysis: Nokia and Marvell partnership to develop 5G RAN silicon technology + other Nokia moves
China gaining on U.S. in AI technology arms race- silicon, models and research
Marvell shrinking share of the RAN custom silicon market & acquisition of XConn Technologies for AI data center connectivity
Intel FlexRAN™ gets boost from AT&T; faces competition from Marvel, Qualcomm, and EdgeQ for Open RAN silicon
China’s telecom industry rapid growth in 2025 eludes Nokia and Ericsson as sales collapse
According to a Chinese government update, “Telecommunications business volume and revenue grew steadily, mobile internet access traffic maintained rapid growth, and the construction of network infrastructure such as 5G, gigabit optical networks, and the Internet of Things was further promoted.”
Figure 1. Cumulative growth rate of telecommunications service revenue and total telecommunications service volume
There were 4.83 million 5G base stations in service in China at the end of November 2025, an increase of 579,000 since late 2024 and 37.4% of the total number of mobile base stations in China. In one year, China claims to have added more 5G base stations than Europe has installed since the 5G technology was first put into service.
The total number of mobile phone users of the top four Chinese telcos (China Mobile, China Telecom, China Unicom, China Broadcasting Network) reached 1.828 billion, a net increase of 38.54 million from the end of last year. Among them, 5G mobile phone users reached 1.193 billion, a net increase of 179 million from the end of last year, accounting for 65.3% of all mobile phone users.
Meanwhile, the total number of fixed broadband internet access users of the three state owned telecom operators (China Mobile, China Telecom and China Unicom) reached 697 million, a net increase of 27.12 million from the end of last year. Among them, fixed broadband internet access users with access speeds of 100Mbps and above reached 664 million, accounting for 95.2% of the total users; fixed broadband internet access users with access speeds of 1000Mbps and above reached 239 million, a net increase of 32.52 million from the end of last year, accounting for 34.3% of the total users, an increase of 3.4 percentage points from the end of last year.
The construction of gigabit fiber optic broadband networks continues to advance. As of the end of November, the number of broadband internet access ports nationwide reached 1.25 billion, a net increase of 48.11 million compared to the end of last year. Among them, fiber optic access (FTTH/O) ports reached 1.21 billion, a net increase of 49.42 million compared to the end of last year, accounting for 96.8% of all broadband internet access ports. As of the end of November, the number of 10G PON ports with gigabit network service capabilities reached 31.34 million, a net increase of 3.133 million compared to the end of last year.
The penetration rate of gigabit and 5G users continued to increase across all regions. As of the end of November, the penetration rates of fixed broadband access users with speeds of 1000Mbps and above in the eastern, central, western, and northeastern regions were 34.6%, 33.8%, 35.8%, and 28.5%, respectively, representing increases of 3.4, 2.6, 4.1, and 4.9 percentage points compared to the end of last year; the penetration rates of 5G mobile phone users were 64.9%, 65.9%, 65.1%, and 65.9%, respectively, representing increases of 8.2, 8.7, 8.8, and 9.6 percentage points compared to the end of last year.
…………………………………………………………………………………………………………………………………………………………………
Separately, Light Reading reports that Ericsson and Nokia sales of networking equipment to China have collapsed.
Ericsson recently published earnings release for the final quarter of 2025 puts China revenues at just 3% of total sales last year. This would equate to revenues of 7.1 billion Swedish kronor (US$798 million). Based on a rounding range of 2.5% to 3.4%, it works out to be between SEK5.92 billion ($665 million) and SEK8.05 billion ($905 million) – down sharply compared with the SEK10.2 billion ($1.15 billion) Ericsson made in 2024, according to that year’s Ericsson annual report.
Nokia does not break out details of revenues from mainland China, instead lumping them together with the sales it generates in neighboring Hong Kong and Taiwan. But this “Greater China” business is in decline. Total annual revenues – which include Nokia’s sales of fixed, Internet Protocol and optical network products, as well as 5G – slumped from almost €2.2 billion ($2.6 billion) in 2019 to around €1.5 billion ($1.8 billion) in 2020, before creeping back up to nearly €1.6 billion ($1.9 billion) by 2022. Two years later, they had fallen to about €1.1 billion ($1.3 billion).
Bar Chart Credit: Light Reading
Nokia has recently indicated the complete disappearance of its China business. “Western suppliers, which is only us and Ericsson, have 3% market share now in China and it’s been coming down, and we are going to be excluded from China for national security reasons,” said Tommi Uitto, the former president of Nokia’s mobile networks business group, at a September press conference in Finland also attended by Justin Hotard, Nokia’s CEO. It implies China’s government is now treating the Nordic vendors in the same way Europe and the U.S. are banning Huawei and ZTE networking equipment.
Nokia revealed in its latest earnings update that Greater China revenues for 2025 had fallen by another 19%, to €913 million ($1.08 billion) – just 42% of what Nokia earned in the region seven years earlier. In the last few years, moreover, Nokia has cut more jobs in Greater China than in any other single region. While figures are not yet available for 2025, the Greater China headcount numbered 8,700 employees in 2024, down from 15,700 in 2019.
Ericsson has significantly reduced its China operations following greatly reduced 5G market share. In September 2021, the company consolidated three operator-specific customer units into a unified structure, impacting several hundred sales and delivery roles within its ~10,000-person local workforce. This followed the divestment of a Nanjing-based R&D center (approx. 650 employees), aligning with strategic pivots away from legacy 2G-4G technologies. The company’s total workforce in Northeast Asia plummeted from about 14,000 in mid-2021 to roughly 9,500 at the end of last year, according to Ericsson’s financial statements.
Exclusion from China would leave Ericsson and Nokia on the outside of the world’s most promising 6G market in 2030. That would intensify concern about a bifurcation of 6G into Western and Chinese variants of IMT 20230 RIT/SRIT standard and the 3GPP specified 6G core network.
References:
https://www.miit.gov.cn/gxsj/tjfx/txy/art/2025/art_7514154ec01c42ecbcb76057464652e4.html
https://www.lightreading.com/5g/ericsson-and-nokia-see-their-sales-in-china-fall-off-a-cliff
China’s open source AI models to capture a larger share of 2026 global AI market
Goldman Sachs: Big 3 China telecom operators are the biggest beneficiaries of China’s AI boom via DeepSeek models; China Mobile’s ‘AI+NETWORK’ strategy
China Telecom’s 2025 priorities: cloud based AI smartphones (?), 5G new calling (GSMA), and satellite-to-phone services
China ITU filing to put ~200K satellites in low earth orbit while FCC authorizes 7.5K additional Starlink LEO satellites
China gaining on U.S. in AI technology arms race- silicon, models and research
SoftBank and Ericsson-Japan achieve 24% 5G throughput improvement using AI-optimized Massive MIMO
SoftBank Corp. and Ericsson Japan K.K. have announced a successful demonstration and deployment of an AI-powered, externally controlled optimization system for Massive MIMO, resulting in a 24% improvement in 5G downlink throughput, increasing speeds from 76.9 Mbps to 95.5 Mbps during periods of high traffic fluctuation.
- Dynamic Beam Patterns: The system automatically adjusts horizontal and vertical beam patterns every minute based on real-time user distribution.
- Packet Stalling Mitigation: By reacting to sudden traffic surges (e.g., during fireworks or concerts), the AI helps prevent “packet stalling,” where data transmission typically freezes due to congestion.
- Commercial Deployment: Following the successful trials at Expo 2025, SoftBank and Ericsson have begun deploying this AI-based system at other large-scale event venues, including major arenas and dome-type facilities in the Tokyo metropolitan area, to manage heavily fluctuating traffic patterns.
Overview of the System:
- An external control device (server) uses user distribution data collected from base stations at one-minute intervals to automatically determine event occurrence using AI
- Dynamically and automatically optimizes the horizontal and vertical coverage patterns of Massive MIMO base stations
Overview of demonstration at Expo 2025:
An AI model was constructed using performance results obtained when multiple coverage patterns were changed in advance as training data. Based on user distribution-related data such as Massive MIMO beam estimation information*2 acquired by an external control device from base stations at one-minute intervals, the system automatically determined event occurrence status and switched base station coverage patterns to optimal configurations.
……………………………………………………………………………………………………………………………….
ABOUT SOFTBANK:
Guided by the SoftBank Group’s corporate philosophy, “Information Revolution – Happiness for everyone,” SoftBank Corp. (TOKYO: 9434) operates telecommunications and IT businesses in Japan and globally. Building on its strong business foundation, SoftBank Corp. is expanding into non-telecom fields in line with its “Beyond Carrier” growth strategy while further growing its telecom business by harnessing the power of 5G/6G, IoT, Digital Twin and Non-Terrestrial Network (NTN) solutions, including High Altitude Platform Station (HAPS)-based stratospheric telecommunications. While constructing AI data centers and developing homegrown LLMs specialized for the Japanese language, SoftBank is integrating AI with radio access networks (AI-RAN), with the aim of becoming a provider of next-generation social infrastructure. To learn more, please visit https://www.softbank.jp/en/corp/
………………………………………………………………………………………………………………………………….
References:
SoftBank’s Transformer AI model boosts 5G AI-RAN uplink throughput by 30%, compared to a baseline model without AI
Softbank developing autonomous AI agents; an AI model that can predict and capture human cognition
Nokia, BT Group & Qualcomm achieve enhanced 5G SA downlink speeds using 5G Carrier Aggregation with 5 Component Carriers
T‑Mobile achieves record 5G Uplink speed with 5G NR Dual Connectivity
Huawei, Qualcomm, Samsung, and Ericsson Leading Patent Race in $15 Billion 5G Licensing Market
Ericsson announces capability for 5G Advanced location based services in Q1-2026
Highlights of Ericsson’s Mobility Report – November 2025
Ericsson integrates Agentic AI into its NetCloud platform for self healing and autonomous 5G private networks
Ericsson announces capability for 5G Advanced location based services in Q1-2026
Ericsson’s 5G Advanced location based services (LBS) offering is a comprehensive suite of innovations designed to redefine location-based services across commercial 5G Standalone (SA) networks. Set for release in Q1 2026, it makes Ericsson the leader in 5G positioning technology, offering a scalable and fully integrated solution on top of Ericsson’s dual-mode 5G Core network.
By embedding positioning as a core 5G SA network capability, Ericsson 5G Advanced location services enables Communications Service Providers (CSPs) to monetize precise location services and expand beyond traditional mobile offerings into verticals such as manufacturing, healthcare, public safety, automotive, drones, and more.
Key benefits:
- High Accuracy: Down to sub-meter for indoor and sub-10 cm for outdoor positioning, enabling precise tracking
- Scalability: Scalable, precise positioning for outdoor applications (automotive, agriculture, drones)
- Seamless Indoor/Outdoor Coverage: Unified 5G positioning technology for both environments.
- Developer & Device Friendliness: No need for device-side apps; improved battery life compared to satellite-based solutions
- Support for Large-Scale Use Cases: Enables massive geofencing, population density analysis, and tracking use cases.
Monica Zethzon, Head of Core Networks, Ericsson, says: “With the launch of 5G Advanced Location Services we are evolving the value of 5G Standalone networks. This innovation gives CSPs the precision and scalability to create differentiated services based on location capabilities.”
Caroline Gabriel, Partner at Analysys Mason, says: “Ericsson’s integrated approach to indoor and outdoor positioning sets a new benchmark in the industry. It addresses critical pain points for operators and enterprises, particularly in sectors where location accuracy is mission-critical.”
The global market for 5G positioning is in its early stages but poised for rapid growth, driven by demand for enhanced precision in diverse sectors. Ericsson’s solution responds to this demand with scalable, developer-friendly capabilities that improve device battery life compared to legacy systems.
This launch further strengthens Ericsson’s location solutions based on Real-Time Kinematics technology, with related devices from Ericsson planned for Q1 2026.
Photo Credit: Ericsson
………………………………………………………………………………………………………………………………………………………….
- Integrated Positioning: Positioning is built into the 5G Standalone (SA) architecture, moving beyond traditional GPS reliance.
- High Accuracy & Efficiency: New techniques improve accuracy (e.g., bandwidth aggregation, carrier-phase measurements) and reduce power consumption for devices.
- AI/ML Integration: Artificial Intelligence/Machine Learning is applied to enhance positioning accuracy, especially for challenging scenarios like beyond-visual-line-of-sight (BVLOS).
- Support for New Devices/Apps: Enables precise tracking for wearables, industrial sensors (RedCap), augmented reality (AR), drone control, and smart grids.
- Beyond-Line-of-Sight (BVLOS): Focus on reliable positioning for industrial and public safety applications where line-of-sight isn’t guaranteed.
- Reduced Power: Solutions target lower power usage, crucial for IoT devices.
- Release 18 (5G Advanced Start): Finalized mid-2024, introduced major LBS enhancements, including RedCap positioning, bandwidth aggregation, and carrier-phase support.
- Release 19 (Ongoing): Continues the evolution, extending LTM (L1/L2-triggered Mobility) and further exploring AI/ML for mobility and positioning.
- Release 20 & Beyond: Will build on these foundations, further evolving towards 6G capabilities.
References:
https://www.ericsson.com/en/press-releases/2026/1/5g-advanced-location-services
5G Advanced offers opportunities for new revenue streams; 3GPP specs for 5G FWA?
What is 5G Advanced and is it ready for deployment any time soon?
Hutchison Telecom is deploying 5G-Advanced in Hong Kong without 5G-A endpoints
China Mobile & ZTE use digital twin technology with 5G-Advanced on high-speed railway in China
Huawei pushes 5.5G (aka 5G Advanced) but there are no completed 3GPP specs or ITU-R standards!
ZTE and China Telecom unveil 5G-Advanced solution for B2B and B2C services
Comparing AI Native mode in 6G (IMT 2030) vs AI Overlay/Add-On status in 5G (IMT 2020)
Executive Summary:
AI integration in 6G specifications (3GPP) and standards (ITU-R IMT 2030) highlights a strategic shift in the telecom industry towards AI-native networks, with telecom industry heavyweights like Huawei, Samsung, Ericsson, and Nokia actively developing foundational technologies. Unlike 5G, where AI and machine learning were limited applications or add-on features over existing architecture, 6G will incorporate AI from the onset with an “AI native” approach where intelligence will allow the network to be smart, agile, and able to learn and adapt according to changing network dynamics.
This transformation is necessary because future 6G networks will be too complex for human operators to manage, requiring AI-empowered and learning-driven networks that can facilitate zero-touch network management through capabilities including learning, reasoning, and decision-making.
- AI-Native Networks: The industry consensus is that 6G will be “AI-native,” meaning artificial intelligence will be built directly into the core functions of network control, resource management, and service orchestration. This moves AI from an optimization layer in 5G to an foundational element in 6G.
AI Native Image Courtesy of Ericsson
…………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………
- Company Initiatives:
- Huawei is focused on making AI a native element of the network architecture (AI-native 6G) rather than an overlay technology, integrating communication, sensing, computing, and intelligence. This vision, called “Connected Intelligence,” involves two aspects: AI for 6G (network automation) and 6G for AI (AI as a Service, AIaaS). More in Huawei Research Areas below.
- Samsung is a major proponent of AI-RAN (Radio Access Network) technology. The company hosted a summit in November 2025 to showcase working AI-RAN technology that autonomously optimizes network performance and is conducting joint research with SK Telecom (SKT) on AI-supported RAN. Samsung sees vRAN (virtualized RAN) as a key enabler for “AI-native, 6G-ready networks”.
- Ericsson emphasizes the necessity of a strong 5G Standalone (5G SA) foundation for an AI future, using AI to manage and automate current networks in preparation for 6G’s demands. Ericsson is also integrating agentic AI into its platforms for more autonomous network management.
- Nokia is deepening its AI push, licensing software to expand AI use in mobile networks and preparing for early field trials in 2026 by porting baseband software to platforms like NVIDIA’s, which opens the door for more advanced AI use cases.
- Industry Analysis and Trends:
- Standardization: 2026 is crucial as formal 6G specification work begins in earnest within 3GPP with Release 21. In WP5D, the IMT 2030 RIT/SRIT standardization work will commence at the February 2027 meeting with the final deadline for submissions at the February 2029 meeting. More in the ITU-R WP5D section below.
- The AI-RAN Alliance is an industry initiative (not a traditional SDO) focused on accelerating real-world AI applications and integration within the RAN. It works alongside SDOs, providing industry insights and pushing for rapid validation and testing of AI-RAN technologies, with a specific focus on leveraging accelerated computing.
- Automation and Efficiency: AI-native algorithms in 6G are expected to deliver extreme spectrum and energy efficiency, significantly reducing operational costs for telcos while improving reliability and performance.
- Monetization Challenges: Despite the technological promise, analysts caution that 6G remains largely theoretical for now. Some operators are stalling on full 5G SA deployment, waiting to move to 6G-ready cores later in the decade, leading to concerns that 5G SA might become an “odd generation.”
- Infrastructure Constraints: The physical demands of AI infrastructure, particularly energy consumption and construction timelines, are becoming operational realities that may bound the pace of AI growth in 2026, regardless of software advancements.
- ITU-R Working Party (WP) 5D is making AI a native and foundational element of the 6G (IMT-2030) system, rather than the “add-on” or “overlay” status it had in 5G (IMT 2020). This shift is being achieved through the definition of specific AI capabilities and requirements that future 6G technologies must inherently support. In particular:
- Defining AI as a Core Capability: The Recommendation ITU-R M.2160 (“Framework and overall objectives of the future development of IMT for 2030 and Beyond”) officially defines “Artificial Intelligence and Communication” as one of the six major usage scenarios and an overarching design principle for IMT-2030.
- Integrating AI into the Radio Interface: WP 5D is actively developing technical performance requirements (TPRs) and evaluation criteria for proposed 6G radio interface technologies (RITs) that inherently incorporate AI/Machine Learning (ML). This includes work on:
- AI-enabled air interface design: This involves the physical layer, potentially moving towards AI-native physical (PHY) layers that can dynamically adapt waveforms and network parameters in real-time, rather than relying on predefined, static configurations.
- AI-driven resource management: AI/ML algorithms will be crucial for real-time optimization of spectral and energy efficiency, managing complex traffic, and ensuring Quality of Service (QoS).
- Enabling AI-Driven Services: The framework for IMT-2030 is designed to support the full lifecycle of AI components, from data collection and model training to deployment and performance monitoring, enabling new AI-driven services and applications directly within the network infrastructure.
- Establishing a Formal Timeline: WP 5D has established a clear timeline for 6G standardization, with specific stages for vision, requirements, evaluation methodology, and specifications. This structured approach ensures that all proposed RITs/SRITs are evaluated against the new AI-native requirements, promoting global alignment and preventing AI from becoming a fragmented, proprietary solution.
- Stage 1 (Vision): Completed in June 2023.
- Stage 2 (Requirements & Evaluation): Targeted for completion in 2026.
- Stage 3 (Specifications): Expected by the end of 2030.
- Purpose: AI is integral to the entire network lifecycle, from initial design and deployment to autonomous operation and service creation.
- Integration Level: Intelligence is embedded across all layers of the network stack, including the physical layer (air interface), control plane, and data plane.
- Scope: AI enables core functionalities such as real-time self-optimization, self- healing capabilities, and dynamic resource allocation, rather than static, predefined configurations.
- Outcome: The creation of a fully cognitive, self-managing, and highly adaptable “intelligence fabric” capable of supporting advanced use cases like real-time holographic communication, digital twins, and autonomous systems with ultra-low latency.
| Feature | 5G (IMT-2020) | 6G (IMT-2030) |
|---|---|---|
| AI Role | Optimization tool (overlay) | Foundational and native element |
| Network Operation | Manual configuration with AI assistance | Autonomous and self-managing |
| Air Interface | Human-designed with some ML optimization | AI/ML-designed and managed |
| Complexity Management | Relies on standard protocols | Manages complexity through embedded AI/ML |
| Services Supported | Enhanced mobile broadband, basic IoT | Integrated AI & Communication, sensing, holographic comms |
–>By embedding AI into the fundamental design principles and technical requirements of IMT-2030, ITU-R WP 5D is ensuring that 6G is an AI-native network capable of self-management, self-optimization, and supporting a vast ecosystem of AI applications, a significant shift from the supplementary role AI played in 5G.
- Agentic-AI Core (A-Core): Huawei unveiled a blueprint for a 6G core network (which will be specified by 3GPP and NOT ITU) where services are managed by specialized AI agents using a large-scale network AI model called “NetGPT”. This allows the network to program, update, and execute its own control procedures automatically without human intervention, based on natural language instructions.
- Network Architecture Redesign: Huawei proposes the NET4AI system architecture, a service-oriented design that moves beyond the 5G service-based architecture. It introduces a dedicated data plane (DP) to handle the massive volume of data generated by AI and sensing services, enabling flexible and efficient many-to-many data flow for distributed learning and inference.
- Integrated Sensing and Communication (ISAC): A core pillar of Huawei’s 6G work is the native integration of sensing with communication. This allows the network to use radio waves for high-resolution sensing, localization, and imaging, creating a “digital twin” of the physical world. The large volume of data collected from sensing then serves as a source for AI model training and real-time environmental monitoring.
- Distributed Machine Learning: Huawei researches deep-edge architecture to enable massive, distributed, and collaborative machine learning (ML). This includes the development of frameworks like a two-level learning architecture that combines federated learning (FL) and split learning (SL) to optimize computing resources and ensure data privacy by keeping raw data local to devices.
- AI as a Service (AIaaS): The 6G network is designed to provide AI capabilities as a service, allowing the training and inference of large AI models to be distributed across the network (edge and cloud). This offers low-latency performance and access to rich data for AI-driven applications like collaborative robotics and autonomous driving.
- Energy Efficiency and Sustainability: The company is researching how native AI capabilities can improve overall energy efficiency by up to 100 times compared to 5G. This involves smart energy control, dynamic resource scaling, and optimizing communication paths for lower power consumption.
- Standardization and White Papers: Huawei is actively contributing to global 6G discussions and standardization bodies like the ITU-R, sharing its vision through publications such as the book 6G: The Next Horizon – From Connected People and Things to Connected Intelligence and various technical white papers. The goal is to define the technical specifications and use cases for 6G that will drive industry-wide innovation by around 2030.
References:
https://www.ericsson.com/en/reports-and-papers/white-papers/ai-native
Roles of 3GPP and ITU-R WP 5D in the IMT 2030/6G standards process
AI wireless and fiber optic network technologies; IMT 2030 “native AI” concept
ITU-R WP5D IMT 2030 Submission & Evaluation Guidelines vs 6G specs in 3GPP Release 20 & 21
ITU-R WP 5D Timeline for submission, evaluation process & consensus building for IMT-2030 (6G) RITs/SRITs
ITU-R WP 5D reports on: IMT-2030 (“6G”) Minimum Technology Performance Requirements; Evaluation Criteria & Methodology
AI wireless and fiber optic network technologies; IMT 2030 “native AI” concept
Highlights of 3GPP Stage 1 Workshop on IMT 2030 (6G) Use Cases
Should Peak Data Rates be specified for 5G (IMT 2020) and 6G (IMT 2030) networks?
GSMA Vision 2040 study identifies spectrum needs during the peak 6G era of 2035–2040
Highlights and Summary of the 2025 Brooklyn 6G Summit
NGMN: 6G Key Messages from a network operator point of view
Nokia and Rohde & Schwarz collaborate on AI-powered 6G receiver years before IMT 2030 RIT submissions to ITU-R WP5D
Verizon’s 6G Innovation Forum joins a crowded list of 6G efforts that may conflict with 3GPP and ITU-R IMT-2030 work
Nokia Bell Labs and KDDI Research partner for 6G energy efficiency and network resiliency
Deutsche Telekom: successful completion of the 6G-TakeOff project with “3D networks”
Market research firms Omdia and Dell’Oro: impact of 6G and AI investments on telcos
Qualcomm CEO: expect “pre-commercial” 6G devices by 2028
Ericsson and e& (UAE) sign MoU for 6G collaboration vs ITU-R IMT-2030 framework
KT and LG Electronics to cooperate on 6G technologies and standards, especially full-duplex communications
Highlights of Nokia’s Smart Factory in Oulu, Finland for 5G and 6G innovation
Nokia sees new types of 6G connected devices facilitated by a “3 layer technology stack”
Rakuten Symphony exec: “5G is a failure; breaking the bank; to the extent 6G may not be affordable”
India’s TRAI releases Recommendations on use of Tera Hertz Spectrum for 6G
New ITU report in progress: Technical feasibility of IMT in bands above 100 GHz (92 GHz and 400 GHz)
Highlights of Ericsson’s Mobility Report – November 2025
The latest issue of the Ericsson Mobility Report states that 5G subscriptions now account for one-third of total mobile subscriptions. Mobile network data traffic grew slightly more than expected – 20 percent between Q3 2024 and Q3 2025. As 5G evolves, service providers are increasingly exploring innovative use cases and new monetization opportunities such as offering differentiated connectivity services and modernizing enterprise IT with 5G.
After many years of hype, network slicing, which requires a 5G SA core network, is finally gaining market traction with 33 communications service providers now offering variations of the technology. Of the 118 network slicing cases discovered by Ericsson’s researchers, 65 have moved beyond proof of concept and into commercial services, either as standalone subscription services or as add-on packages for consumer or business customers. Ericsson attributes this growth spurt to more widespread deployment of 5G SA core networks.
Looking further ahead, the 6G RAN standardization process has begun in 3GPP and ITU-R WP5D, with the first commercial launches expected in front-runner markets.
–>However, there has been no work initiated on the 6G core network in either 3GPP or ItU-T.
Ericsson’s report says the U.S., Japan, South Korea, China, India and some Gulf Cooperation Council countries are the 6G leaders. Global 6G subscriptions are likely to reach 180 million by the end of 2031, the report predicts.
We think that forecast is highly unlikely as the IMT 2030 (6G) RIT/SRITs recommendation won’t be completed till the end of 2030 with initial deployments sometime in 2031.
…………………………………………………………………………………………………………….
Data from Omdia, a Light Reading sister company, shows Ericsson, Huawei and Nokia were even more dominant last year than they were in 2023, growing their combined market share by 2.3 percentage points over this period, to 77.4%. Besides China’s ZTE, the only other contender with more than a percentage point of market share was Samsung.
References:
https://www.ericsson.com/en/reports-and-papers/mobility-report/reports/november-2025
Dell’Oro: 4G and 5G FWA revenue grew 7% in 2024; MRFR: FWA worth $182.27B by 2032
Ericsson’s revenue drops, profits soar; deal with Vodafone and partnership with Export Development Canada look promising
Latest Ericsson Mobility Report talks up 5G SA networks and FWA
Ericsson Mobility Report touts “5G SA opportunities”
Ericsson Mobility Report: 5G monetization depends on network performance
Ericsson Mobility Report: 5G subscriptions in Q2 2022 are 690 million (vs. 8.3 billion total mobile users)
Ericsson’s revenue drops, profits soar; deal with Vodafone and partnership with Export Development Canada look promising
Ericsson’s 3rd quarter 2025 results released today showed a 9% drop in revenues, to 56.2 billion Swedish kronor (US$5.9 billion), compared with the same period last year. Ericsson’s gross margin rose 2% to 47.6%. U.S. sales fell by as much as 17% year-over-year for the 3rd quarter, to about SEK22.5 billion ($2.4 billion), after an especially busy period in 2024. And the only region where Ericsson realized any growth was northeast Asia, and that was due to Japan’s new 5G rollout.
At Ericsson’s big mobile networks unit, sales fell 11% year-over-year, to SEK35.4 billion ($3.7 billion), while the decline on a constant-currency basis was just 4%. The division’s operating income also slid by 6%, to SEK7.1 billion ($740 million).
Sales were much better at the company’s cloud software and services group, responsible for the development of Ericsson’s core network software as well as its business and operational support systems. Reported sales rose 3%, to SEK15.3 billion ($1.6 billion), while Ericsson put the organic improvement at 9%. More importantly, it swung from an operating loss of SEK400 million ($42 million) a year earlier to a profit of SEK1.7 billion ($180 million).
Net income soared by an astonishing 191%, to SEK11.3 billion ($1.2 billion). That sharp increase in net income was due to Ericsson’s recent sale of iconectiv, a provider of number-portability and data-exchange services, to a private equity firm. The deal landed Ericsson a capital gain of SEK7.6 billion ($800 million) that flattered its profits at the operating income level. In Stockholm, Ericsson’s share price soared more than 14% in mid-morning trade, although it remained almost 2% below its level at the start of the year.
CEO Börje Ekholm said on today’s earnings call: “The margin expansion reflects actions we’ve taken over the last years to increase operational excellence and efficiency, including the work we’ve done on our cost base. Over the last year, we’ve reduced our headcount by some 6,000, leveraging new ways of working, and that of course includes AI.”
Since the end of 2022, the year Ericsson acquired VoIP software developer Vonage for $6.2 billion, headcount has fallen by more than 15,600, to just 89,898 at the end of June, the company revealed in its latest earnings report.
The Vonage business suffered a 17% drop in sales, to SEK3.2 billion ($330 million), and saw its loss widen by 50%, to SEK600 million ($63 million). It is where Ericsson believes it can monetize the network application programming interfaces (APIs) that will link software apps to networks and hopefully revitalize the 5G market. However, that’s not happening yet.
“The geopolitical situation has required us to shift resources a bit politically. As we went through that transition, we duplicated a large part of the R&D spend. We don’t need to have that anymore as we have relocated R&D,” said Ekholm. “We are not going to jeopardize technology leadership and if we feel there is any risk – and that is a risk I don’t see today – then we would of course need to reassess.”
After years of growth, R&D spending fell by 10% year-over-year for the first nine months of 2025, to SEK35.8 billion ($3.8 billion), prompting concern among analysts that Ericsson could lose competitiveness versus Chinese rivals.
AI is now being used to refine the algorithms that are fed into Ericsson’s software products, said Per Narvinger, the head of Ericsson’s mobile networks business group, on a call with Light Reading. No indication was given if that would reduce headcount any further.
Ericsson hopes the new 5G contract it announced with Vodafone earlier today will boost sales in Europe, where underinvestment in midband 5G coverage and the “standalone” variant of 5G have been constant bugbears for the company. After the rollout of “non-standalone” 5G, which maintains the 4G core, operators just continued to sell a “4G plus” service, Ekholm said.
“It was the established business model of most operators around the world, so it became very natural to take that step and then use 5G almost as a marketing icon on the phone, but, in reality, it didn’t give the extra capabilities,” he added. Standalone features such as low latency and network slicing will be critical in future apps, Ekholm correctly said, arguing that 6G will necessitate edge cloud and AI investments that have also not yet happened.
In summing up, Ericsson said “Increased uncertainty remains on the outlook, both in terms of potential for further tariff changes as well as in the broader macroeconomic environment.”
Looking ahead:
– Continue to invest in technology leadership to strengthen competitive position
– Future-proofed Open RAN-ready portfolio
– New use cases to monetize network investments taking shape
● AI applications becoming a key driver for network investments
● Structurally improving the business through rigorous cost management
……………………………………………………………………………………………………………………………………………………….
Separately,
Ericsson today announced the signing of a USD $3 billion partnership agreement with Export Development Canada (EDC) to expand investment in Canadian research and development, deepen domestic supply chains, and accelerate next-generation technologies including 5G, Cloud RAN, AI, and quantum innovation.
Börje Ekholm, President and CEO, Ericsson, says: “Canada is one of Ericsson’s most important hubs for global research and development, and this partnership with Export Development Canada will allow us to scale that leadership even further. By strengthening our collaboration with Canadian businesses, universities and government partners, we can accelerate breakthroughs in 5G, quantum, and Cloud RAN that will drive growth, create opportunities, and reinforce Canada’s position as a global leader in next generation networks.”
With more than 3,100 employees nationwide and R&D centres in Ottawa, Montreal, and Toronto, Ericsson Canada is at the heart of the company’s global innovation footprint. Canadian teams are driving advancements in 5G, 5G Advanced, and 6G, while also contributing to new research in quantum communications and AI-powered network management.
The three-year partnership will enable Ericsson to expand its Canadian-led innovation and global projects with the support of financial and insurance solutions from EDC. By reinforcing Ericsson’s Canadian supply chain and connecting the company with innovative domestic businesses, the agreement will also amplify Ericsson’s ability to bring Canadian technology to the world, strengthen competitiveness, and create new opportunities for Canadian companies within Ericsson’s global network of partners.
Across all wireless network equipment vendors, annual sales of RAN products fell from $45 billion in 2022 to $35 billion last year, according to Omdia, a Light Reading sister company. Market research firms Omdia and Dell’Oro have encouragingly guided for a more stable market this year.
Most wireless network providers have seen no incentive to spend more on 5G when their returns to date have been so disappointing. And there is skepticism about the business case for low latency services and network slicing. Telcos increasingly sell large bundles of gigabytes to their customers and have struggled to monetize other features.
References:
https://www.lightreading.com/5g/ericsson-says-world-is-flat-amid-us-gloom-and-keeps-cutting
https://www.ericsson.com/en/press-releases/6/2025/ericsson-edc-advance-canadas-technology-leadership
Ericsson integrates agentic AI into its NetCloud platform for self healing and autonomous 5G private networks
Ericsson CEO’s strong statements on 5G SA, WRC 27, and AI in networks
Ericsson completes Aduna joint venture with 12 telcos to drive network API adoption
Ericsson reports ~flat 2Q-2025 results; sees potential for 5G SA and AI to drive growth
Ericsson revamps its OSS/BSS with AI using Amazon Bedrock as a foundation
Ericsson’s sales rose for the first time in 8 quarters; mobile networks need an AI boost
Beyon partners with Ericsson to build energy-efficient wireless networks in Bahrain
Latest Ericsson Mobility Report talks up 5G SA networks and FWA
Ericsson and e& (UAE) sign MoU for 6G collaboration vs ITU-R IMT-2030 framework
