Edge-Native: The new paradigm for operating and developing Edge Apps
summary
Claims that cloudlike devops work at the edge by treating location as a scaling dimension, not just processes, and that many edge use cases now include AI/ML (roughly 50 percent of conversations). Describes a deployable control plane built on lightweight Kubernetes and an experimental platform called grey bear that boots a Raspberry Pi, claims fast site deployment and a demo deploying event driven display to three sites. Shows model splitting and federated learning with an open framework flame, increasing FPS from ~1.1 to ~1.7 in an experiment.
AI-written summary · about
chapters (5)
- 0:00Start
- 3:02Can We Make the Edge Feel like the Cloud from a Development and Operations Perspective
- 8:09Concurrency and Scale
- 27:33Federated Learning
- 27:52What Is Federated Learning
related talks
19mFrom Multicloud to Edge Native: Distributed, intelligent & secure
25mAccelerating Data Processing and Consumption for Building Modern Apps at the Edge
10mThe Rise and Impact of Edge AI
22mEdge AI Use Cases in IoT — Unlocking Intelligence at the Network Edge
29mEdge Intelligence Revolution
21mA New Architecture for the Intelligent Edge
09mWelcome + Opening Remarks, Program Committee Members
24mDemo of Kubernetes at the Edge for AI Model Deployments