edge-native

Definition

Designed from the outset to run on the device or a nearby gateway rather than in a data centre. The contrast is with cloud models retrofitted for the edge by compression: edge-native means memory ceilings, power budgets, and intermittent connectivity were assumptions during design, not problems solved afterwards. The payoff is low latency and data that never leaves the device; the price is a hard cap on model size and compute.

Explainers using this term

• Paper Explained: FreeToken — Treating Your Own PC as a Single Elastic Inference PlatformFreeToken: Efficient Edge-Native MoE Serving with Bandwidth-Adaptive Execution