Edge computing places computation and data processing near the users or devices that generate or consume data. It reduces geographical and last-mile latency by decentralizing workloads across multiple edge locations instead of relying on one central cloud region. section 3.2.2
Near edge uses points of presence inside or near traditional data centers, while far edge moves resources closer to the user’s network, including homes, factories, point-of-sale systems, and IoT environments. Because “edge” covers different infrastructure layers, identify the specific approach before evaluating its latency benefits. section 3.2.2
Regional access and colocated APIs
A user in Tokyo can access a nearby edge location instead of communicating with a central region in Virginia. Similarly, placing a REST API near its data, such as by embedding SQLite in the server, can reduce latency substantially. section 3.2.2 section 3.6
Apply edge computing as a colocation strategy:
- Place computation near the users generating or consuming the data. section 3.2.2
- Place compute and data together when applications repeatedly access a database or another system. section 3.6
- Compare near-edge and far-edge options against the application’s actual latency requirements rather than treating “edge” as a uniform architecture. section 3.2.2
