Using Qt Quick with low end devices (eg i.MX6)
summary
Shows how Qt Quick runs on i.MX6 dual lite and SoloX by giving concrete KPIs and fixes: customers require boot <10s, usable <12s, avg CPU ≤10%, target 30 fps, ≤70 MB RAM, 16 months uptime. Solutions include lazy startup with an early image, modular app design, a hybrid QML caching strategy (keep heavy screens after first load), precreate list delegates offscreen, implement a C++ page cache that parents only two visible screens, avoid full‑screen fragment shaders on Vivante drivers, and use the scene‑graph profiler to find batching issues.
AI-written summary · about
chapters (25)
- 0:00Intro
- 0:35About us
- 1:16Why this talk
- 2:15The iMX6
- 3:24KPIs Expectations
- 6:01Boot Time Optimization
- 6:48User Experience
- 8:12Memory Management
- 8:35Static Memory Management
- 9:09Hybrid Memory Management
- 9:48Delegate Management
- 11:28Free Management
- 11:43Page Heap
- 12:57State Machine
- 13:30Shaders
- 15:29Rendering
- 16:50Performance Issues
- 18:23Conclusion
- 20:36Demo
- 21:28UI
- 22:15Background
- 23:20Drivers
- 24:35Performance degradation
- 25:07Performance segregation
- 25:48fragmentation
related talks
01h 02mCreating Performant Qt Quick Apps
54mCustomizing Qt to create first class graphical experiences on highly customized hardware
45mQt, multimedia, and computer vision – a great match for Embedded Linux
49mScaling Qt from Desktop to Servers and Micro-controllers
01h 08mQt on MeeGo
28mAccelerating Performance of Qt on TI's Embedded Processors
58mQt for Device Creation
27mQt and Qt Quick on the Freebox Player set-top box