Ericsson's RAN Compute Platform Selected by NTT DOCOMO for AI-Ready 5G Networks

The energy efficiency gains directly impact operational costs, with reduced power demands for high-density urban sites ◉ thefastmode.com · 1. Ericsson’s AI in RAN software, deployed alongside the platform, claims 20% higher downlink throughput in live tests ◉ techblog.comsoc.org · 6. This integration positions DOCOMO to leverage AI-driven network automation, aligning with broader industry trends toward intelligent infrastructure ◉ techblog.comsoc.org · 6.

The platform’s modular design reduces hardware sprawl while enabling scalable AI algorithm deployment ◉ techblog.comsoc.org · 8. By prioritizing custom silicon over general-purpose GPUs, Ericsson addresses both performance and energy efficiency concerns ◉ 5gstore.com · 5.

Technical Advancements and Industry Implications

The RAN Compute platform’s integration of AI-native operations aligns with Ericsson’s broader strategy to embed advanced AI capabilities directly into 5G infrastructure. This approach, highlighted in Ericsson’s AI in RAN software launch, enables improved network performance and energy savings without additional hardware investments ◉ telecomlead.com · 3. By leveraging proprietary silicon, the platform reduces reliance on general-purpose GPUs, a move that addresses both computational efficiency and energy consumption challenges in high-density urban deployments ◉ 5gstore.com · 5.

Industry adoption of AI-driven RAN solutions is accelerating, with Ericsson’s platform positioned to support 5G Advanced and future network evolution. The June 2026 Mobility Report underscores the growing role of 5G as "critical, intelligent infrastructure," projecting that 5G will carry 85% of global mobile data traffic by 2031 ◉ techblog.comsoc.org · 7. NTT DOCOMO’s deployment of the RAN Compute platform reflects this trend, enabling the operator to meet rising demand while maintaining energy efficiency gains of over 50% compared to legacy systems ◉ thefastmode.com · 1.

The platform’s modular design also supports scalable AI algorithm deployment, allowing operators to adapt to evolving network requirements without hardware overhauls. This flexibility is critical as operators like DOCOMO explore "In-Network Computing" architectures, where AI processing occurs within the network itself to optimize real-time data handling ◉ docomo.ne.jp · 9. Such advancements position Ericsson to capture a larger share of the AI-RAN market, which is expected to grow rapidly as telcos transition from isolated AI pilots to production-grade automation ◉ techblog.comsoc.org · 6.