Download presentation
Presentation is loading. Please wait.
Published byMatthew McDowell Modified over 6 years ago
1
Optimizing Edge-Cloud IoT Applications for Performance and Cost
Peter Bodik Microsoft Research
2
Can we make IoT scalable and cheap?
Many IoT scenarios that we don’t know how to solve/enable need better sensors, cheaper connectivity, … However, many other verticals solved Problem: simplify development, scale up, and make cheap process store visualize alert
3
Complex deployment of simple IoT apps
Challenges: deployment monitoring cost estimates placement resource allocation Complex deployment of simple IoT apps application: devices stream analytics Azure function DB edge cloud devices stream analytics Azure function DB event hub IoT devices stream analytics IoT hub Azure function DB pass-thru function
4
Many ways to implement/optimize the app
Communication: batching, encoding, compression, approximation, … Processing: Azure Stream Analytics, Function, SQL stored procedure, custom VM, … (Time series data) storage: SQL server, Cosmos DB, Time Series Insights, Tables, Blobs, … Manual and complex optimization
5
End-to-end management
Declarative Spec Deployment Verify Compile Deploy End-to-end management on Edge-Cloud Monitoring Perf. Cost Bottlenecks Placement & allocation Time-series Storage Communication Optimization …
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.