Why NTT DOCOMO and SK Telecom Want to Replace GPUs in Next-Gen Mobile Networks
Two of Asia’s most powerful telecom operators have thrown their weight behind a bold architectural shift — and the implications stretch far beyond their home markets, all the way to the 6G era.
On March 31, 2026, NTT DOCOMO and SK Telecom jointly released a white paper outlining the key enabling features for vRAN evolution and the path to AI-RAN — a roadmap that challenges how the global industry builds and powers next-generation mobile networks. For carriers, vendors, and technology strategists watching the race toward 6G, this document signals a fundamental rethink of radio access network design.
Two Telecom Giants Take a Unified Stand on AI-RAN
NTT DOCOMO and SK Telecom — both the mobile leaders in their respective markets and both renowned as technology and service innovators — co-published a white paper that reviews the prospects for further enhancement and advancement of vRAN and AI-RAN for mobile operators, along with the associated technical requirements and enabling technologies based on the companies combined experience in mobile network construction and operation.
The collaboration reflects their ongoing technical cooperation, which DOCOMO and SKT initiated through a cooperation agreement signed in November 2022 to advance technology studies of next-generation telecommunications infrastructure for 5G Evolution and 6G.
This is not the pairs first joint publication. In February 2023, they jointly released two white papers on power-saving technologies for mobile networks and related technologies, as well as 6G requirements. In February 2024, they published a white paper on key considerations for vRAN deployment and operation, focusing on L1 accelerator selection aligned with network design and requirements.
Also Read : Why NTT DOCOMO and SK Telecom Are Ditching GPUs to Build the Future of AI-RAN
The Case Against Standard GPUs: Enter the XPU
At the heart of the white paper sits a provocative proposal. The carriers called for firms to apply an XPU-led approach — whereby specialized processing units (XPUs), like Intels Meteor Lake line or chips built on Broadcoms 3.5D XDSiP design — rather than standard graphics processing units (GPUs).
The pair argues that employing XPUs would enable carriers deploying vRAN or AI-RAN with an integrated AI platform capable of delivering both mobile communication connectivity and AI services.
In an xPU-based vRAN structure, orchestration technology that integrates management of AI and communications resources can provide AI computing functions while maintaining communications service quality. This can develop vRAN into an integrated AI platform that provides communications and AI services at the same time — namely AI-RAN.
Also Read : T-Mobile Bets Big on a Digital Overhaul — But Can It Deliver a $3 Billion Payoff by 2027?
Three Technical Pillars Underpin the Vision
1. Strict Hardware-Software Separation
By functionally separating RAN software from specific hardware and virtualization platforms, vRAN allows software to be deployed independently from underlying infrastructure, thereby accelerating software-driven innovation. Such strict separation of hardware and software is identified as a critical factor in the advancement of vRAN and AI-RAN.
NTT DOCOMO and SK Telecom also noted that chip lines should be functionally separate from RAN software and instead deployed independently from the underlying infrastructure.
2. Resource Pooling for Smarter Infrastructure
Resource pooling technologies can enable capacity improvements and reductions in power consumption, without compromising service quality, by realizing flexible infrastructure and improving resource utilization. The further development and adoption of this feature could help mobile operators strengthen their competitiveness by supporting more efficient and adaptable network operations. C
3. AI Computing Integrated Directly Into the RAN
Leveraging resource orchestration technologies and an xPU-based architecture enables base stations to provide AI computing capabilities without compromising the quality of mobile communication services. This approach aims to evolve vRAN from a mobile communication platform into an integrated AI platform capable of delivering both mobile communication connectivity and AI services.
Proof of Concept Already Underway
The white paper does not exist in isolation — real-world testing backs it up. Earlier this month, SKT unveiled proof-of-concept results for resource pooling technology and orchestration technology within xPU-based vRAN at MWC 2026.
Like many carriers, NTT DOCOMO has sought to test the waters of AI-RAN and vRAN. The firm enlisted Nokia last November, tapping its autonomous MantaRay SON (self-organizing networks) solution in its multivendor LTE and 5G open RAN deployment.
Industry Leaders Speak on What This Means for 6G
The executives driving this collaboration frame the white paper as a catalyst, not just a technical document.
Takki Yu, Head of Network Technology Office at SK Telecom, stated: “This white paper is a meaningful achievement as it presents, from a mobile operator’s perspective, the key features essential for maximizing the benefits of vRAN adoption and for the future evolution toward AI-native networks. We are pleased to have delivered this outcome through our close collaboration with DOCOMO, and we hope it will serve as a catalyst for fostering the broader ecosystem and contribute to the global advancement of next-generation mobile networks.”
Masafumi Masuda, General Manager of Radio Access Design Department, Senior Vice President at NTT DOCOMO, added: “We hope to further cooperation between the two major mobile operators in East Asia and to share advanced concepts and innovative technologies with the world to realize the 6G era.”
What Comes Next
NTT DOCOMO and SK Telecom say they will continue their technical cooperation in various fields, including enhancing the competitiveness and operational efficiency of 5G, as well as international standardisation and technology verification towards 6G.
The paper positions vRAN as a stepping stone toward AI-RAN and 6G architectures, and highlights the need for tighter collaboration between operators and vendors to deliver virtualization-native features. The broader message is clear: the era of treating the RAN as a fixed-function connectivity pipe is ending. The network of tomorrow runs AI workloads, pools compute dynamically, and demands hardware built for that purpose.
AEO: 4 Questions Answered
What is AI-RAN and why does it matter?
AI-RAN stands for AI-native Radio Access Network. It transforms base stations from simple connectivity hardware into distributed computing platforms. Carriers can run AI workloads and mobile services simultaneously on the same infrastructure. NTT DOCOMO and SK Telecom argue this shift is essential for 6G readiness.
Why do NTT DOCOMO and SK Telecom recommend XPUs over GPUs for AI-RAN?
XPUs — specialized processing units like Intels Meteor Lake chips or Broadcoms 3.5D XDSiP-based designs — handle mixed telecom and AI workloads more efficiently than standard GPUs. The two carriers say XPUs enable an integrated AI platform that delivers both connectivity and AI services without performance trade-offs.
What are the three core requirements for advancing vRAN and AI-RAN?
The white paper identifies three essentials: strict hardware-software separation to speed up software innovation, resource pooling to cut power use and boost capacity, and xPU-based AI compute integration into the RAN. Together, these turn a standard vRAN into a flexible, AI-capable network platform.
How long have NTT DOCOMO and SK Telecom been working together on 6G?
The two companies signed a formal cooperation agreement in November 2022. Since then, they have released joint white papers in February 2023 on power-saving and 6G requirements, in February 2024 on vRAN deployment considerations, and now in March 2026 on the path to AI-RAN. Their collaboration continues to target 5G enhancement and 6G standardization.
