Unverified50% confidenceOpinionExact time
The core judgment of IoD is: for split inference to truly become a network service, service intent, runtime partitioning, and tensor transport must all be simultaneously visible to the mobile data plane.
1
Sources
50%
Confidence
Long-term
Relevance
10/14/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe research proposes Inference on Demand (IoD), attempting to turn split inference from a private implementation at the application layer into a service that can be recognized and scheduled by the network on the 5G user plane.68% similarUnverifiedThere is still a distance from the testbed to live networks, and the deployment density of programmable UPFs, the standardization progress of cross-vendor APIs, and the overhead of inference context under real congestion scenarios are all questions IoD needs to continue answering.67% similarUnverifiedThe key finding of the experiment is: adaptive split selection can run as a network-facing service while avoiding the rigid cost curve of static configuration.65% similarUnverifiedThe hosted model implies dependence on third-party services, which requires careful consideration in data-sensitive scenarios.65% similarUnverifiedThe design adopts a separation of concerns: the Data Space handles governance and compliance, MCP handles protocol translation and tool exposure, and the LLM agent handles intent understanding and task orchestration.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1028679API
curl https://kongchang.com/api/v1/knowledge/claims/1028679MCP
get_claim(id=1028679)