SK Telecom Signs MOU with Arm and Rebellions for AI Infra

Share this article
Share this article
Prioritise Us on Google
Arm's AGI CPU was co-designed by Meta, and SK Telecom plans to deploy it along with Rebellions' upcoming RebelCard in their AI infrastructure buildout (Credit: Arm)
SK Telecom is working with Arm and Rebellions to test AI infrastructure that improves inference efficiency and supports telco network services

SK Telecom has signed a Memorandum of Understanding (MOU) with Arm and Rebellions to develop next-generation AI infrastructure, with a focus on telco network performance and efficiency.

The agreement centres on combining Arm’s AGI CPU with Rebellions’ RebelCard accelerator to support AI inference within AI data centres that underpin telco services.

The companies are planning to test and verify the combined solution inside SK Telecom’s AI data centres.

The facilities support core telco operations, including network optimisation and data processing across mobile and broadband systems.

Youtube Placeholder

Telco networks shape AI infrastructure demand

Telco providers are increasingly depending on AI to manage network traffic while predicting faults and improving service delivery.

This places pressure on infrastructure to handle large-scale inference workloads with efficiency and stability.

The collaboration focuses on a heterogeneous computing structure, which combines a CPU with a dedicated AI accelerator chip.

In this model, the CPU manages general computing tasks such as system control and data handling, while the accelerator focuses on AI inference calculations.

Arm’s AGI CPU is designed for high-density environments while Rebellions’ RebelCard acts as an AI accelerator, built to process inference workloads at scale.

Together, they form an AI-ready alternative to traditional graphics processing unit, or GPU-based systems.

Rebellions' existing Rebel-Quad AI accelerator - the chip that will be used in the agreement will be Rebellions' upcoming RebelCard (Credit: Rebellions)

This development helps to reduce power consumption and operating costs while maintaining processing efficiency.

Lower energy use directly affects network economics for telcos as data centres account for a growing share of operational expenditure.

Eddie Ramirez, Vice President of Go-to-Market for the Cloud AI Business Unit at Arm, says: “The rapid growth of AI inference is driving demand for new data centre infrastructure designed for large scale deployments.

“Partners like SK Telecom and Rebellions played an important role in helping us build the Arm AGI CPU and modernize AI inference infrastructure.”

Eddie Ramirez, Vice President of Go-to-Market, Cloud AI Business Unit at Arm (Credit: Arm)

Testing AI servers for telecom use cases

SK Telecom plans to deploy servers using the combined CPU and accelerator architecture in its AI data centres.

These tests will assess both performance and stability under real-world telco conditions, where latency and uptime are critical for operations.

The company is also reviewing the use of its sovereign AI foundation model, A.X K1, on the new infrastructure, which ensures control over data and compliance with local regulations.

Running this model within a telco environment enables providers to process sensitive data locally while also improving response times for services such as virtual assistants and fraud detection.

Lee Jae-shin, Head of AI Business Development at SK Telecom, says: “By offering a full package that combines infrastructure optimised for inference with our sovereign AI foundation model A.X K1, we will further enhance the competitiveness of our AI data centres.”

Lee Jae-Shin, Head of AI Business Development at SK Telecom (Credit: SK Telecom)

From chips to telco-ready platforms

Arm’s entry into designing its own data centre silicon marks a change in its role within the telco and cloud ecosystem.

The AGI CPU, which was announced in March, targets large-scale AI deployment which aligns with the needs of telco operators building distributed and centralised data infrastructure.

Rebellions RebelCard forms part of a full-stack approach, which includes both hardware and supporting software required to run AI workloads efficiently. 

Oh Jin-wook, Chief Technology Officer of Rebellions, says: “By providing our ‘RebelCard’ – which offers overwhelming performance and power efficiency – alongside our full-stack software, Rebellions has become a core pillar supporting next-generation AI data centres.

“We expect this ‘one-team’ collaboration of experts to serve as a significant precedent in the industry for building AI-specialised infrastructure.”

Oh Jin-wook, Chief Technology Officer at Rebellions (Credit: Rebellions)

The three companies have already demonstrated the potential of this architecture, combining their chips to run an agentic AI service based on OpenAI’s GPT OSS 120B. 

The ability to integrate efficient AI infrastructure directly into data centre environments supports both service delivery and network operations for telcos.

SK Telecom’s collaboration with Arm and Rebellions reflects a move towards tailored infrastructure that aligns with the demands of modern telco networks, with AI playing an increasing role in managing complexity and scale.

Executives