6G Computing Service

Description

The 6G Computing Service is a service provided by the 6G network that utilises computing resources located in the Service Hosting Environment (SHE). It can be consumed by a subscriber through a User Equipment (UE) or by a third party, making compute capability a first-class network service rather than an application-layer afterthought. This concept is defined in 3GPP TR 22.870, the study on 6G use cases and service requirements, and represents one of the foundational service enablers for the sixth generation of mobile networks.

Architecturally, the 6G Computing Service builds on the convergence of communication and computing that began with Multi-access Edge Computing (MEC) and 5G edge computing enablers. The Service Hosting Environment is the logical locus where computing resources — CPU, GPU, storage, accelerators, and associated runtime environments — are hosted and managed. The 6G system exposes these resources through standardized service interfaces so that a UE or an external application provider can request, consume, and manage compute sessions without needing to know the underlying infrastructure topology. The network is responsible for discovering suitable hosting environments, establishing connectivity, and enforcing policies such as location constraints, latency budgets, and data sovereignty.

How it works in practice involves several functional steps. First, a service consumer (UE or third party) requests computing service with certain requirements, such as compute capacity, latency, geographic placement, or affinity to a data source. Second, the 6G system performs compute-aware selection, matching the request to a Service Hosting Environment that satisfies the constraints. Third, connectivity is established between the consumer and the hosting environment, potentially using edge computing traffic influence mechanisms. Fourth, the compute session runs, with the network able to monitor and, if needed, migrate or scale the session. Finally, the session is released and resources are returned to the pool.

Key components include the Service Hosting Environment itself, the compute service exposure function that advertises available resources, the session management functions that bind consumers to compute instances, and the policy and charging functions that govern usage. The service also interacts with network slicing, since a computing service may be delivered as part of a dedicated slice with specific QoS and isolation guarantees.

The 6G Computing Service matters because future applications — immersive XR, AI inference at the edge, digital twins, and autonomous systems — require compute resources to be as dynamically available and as reliably connected as radio resources. By defining computing as a native 6G service, 3GPP enables operators to monetize compute, developers to write portable applications, and users to obtain consistent performance regardless of where the compute physically resides.

Purpose & Motivation

The 6G Computing Service exists to make computing resources a native, standardized offering of the mobile network rather than an external add-on. In 5G, edge computing was enabled through mechanisms such as Edge Application Server discovery and traffic influence, but compute itself was not a first-class service of the network; applications had to discover and manage compute largely on their own. This created fragmentation, limited portability, and made it difficult to guarantee end-to-end performance for compute-intensive workloads.

The motivation for defining a 6G Computing Service is threefold. First, emerging applications such as AI/ML inference, extended reality, and robotics require compute to be placed close to the user with predictable latency, and the network is best positioned to select and manage that placement. Second, operators need new revenue streams beyond connectivity, and offering compute as a service allows them to monetize distributed infrastructure. Third, third-party providers need a uniform way to consume operator-hosted compute without bespoke integrations per operator.

By standardizing the service in TR 22.870, 3GPP provides a common vocabulary and set of requirements that will guide normative work in later releases. It addresses the limitations of previous approaches where compute was invisible to the network, enabling compute-aware mobility, session continuity, and policy enforcement that were not possible when compute lived entirely outside the 3GPP domain.