Loading...
Run computing close to where data is created instead of in a distant central cloud.
Quick Definition
Edge computing processes data near where it is generated, at or close to end users or devices, rather than in a central cloud region. It reduces latency and bandwidth costs for real-time applications, while distributing infrastructure and its associated management overhead.
Edge computing moves processing from large central data centers to locations near the user or device: a factory floor, a retail store, a cell tower, or a small regional site. The goal is speed and resilience. When data is processed a few miles away instead of a thousand, responses arrive in milliseconds and keep working even if the connection to the main cloud drops.
The cost story has two sides. Edge reduces data transfer charges because raw data does not have to travel to a central region for processing. But it adds many small footprints to manage, so cost management and monitoring must stretch across far more locations.
Example. A chain of supermarkets runs checkout fraud checks on a small server inside each store. Decisions take 20 milliseconds instead of 400, and the stores keep selling even during an internet outage.
Edge complements the central cloud rather than replacing it. Heavy training and analytics stay central; fast decisions move to the edge. The Cloud Cost Management Guide explains how to keep visibility when infrastructure spreads out.
A CDN caches and serves content near users. Edge computing runs your actual application logic near users, which covers far more use cases.
It cuts data transfer and latency costs but adds distributed infrastructure to operate. The net effect depends on how data-heavy your workload is.
Workloads that need very low latency or must survive network drops: manufacturing, retail, vehicles, gaming, and real-time video.