The Impact of Micro-Optimizations on Large-Scale Apps
summary
Shows measured wins not theory for Android apps including 50% inflation-time reduction when replacing conditional UI with view stubs, and a measurable drop in frame drops and higher engagement after switching a high-volume grid to a single-row vertical layout. Replacing sequential with parallel asset downloads with capped concurrency speeds rendering but raises thread and memory pressure on low-end devices. Advises measure with sampled production traces and A/B experiments, use percentiles to find outliers, and guard changes with device-specific limits.
AI-written summary · about
chapters (13)
- 0:00Introduction
- 1:06Optimizations and constraints
- 5:42Macro vs micro optimizations
- 8:59Micro optimizations at scale
- 10:44Performance metrics
- 13:57Fine-grained measurement
- 16:43Performance data analysis
- 20:57Case study: Parallel asset downloads
- 22:20Case study: Layout hierarchy
- 26:38Case study: View stub optimizations
- 29:52Case study: Context locks
- 32:23Best practices
- 33:50Conclusion
related talks
46mThe Impact of Micro-Optimizations on Large-Scale Apps
38mLow effort, High impact Android performance wins
42mPerformance Paradox: Navigating the Complexities of Mobile App Optimisation
39mAndroid App Performance in a Nutshell
53mPractical Performance Tuning on Android
29m10 Ways To Improve Your Android App Performance
30mLearn from developers: Lyft and Google Maps share their experiences optimizing app performance
36mPerformance Optimization in Android Applications