Practical Optimizations
summary
Shows concrete wins: profiling with the Jetpack Benchmark Library and inspecting bytecode/assembly with Cutl Explorer finds fixes that actually matter. Example fixes include swapping loop order to cut a Pixel 2 bitmap scan from ~21,200,000 ns → ~4,000,000 ns, replacing check with checkPrecondition 124,000 ns → <100 ns, making functions top‑level to avoid companion init costs, using a scatter map to drop per‑entry overhead to 1 byte and make inserts ~1.7× and iteration ~1.8× faster, and R8/ART/ baseline/profile work that reduces startup tens of milliseconds.
AI-written summary · about
chapters (9)
- 0:00Jetpack Compose optimizations
- 1:11Optimization opportunities
- 7:50Cache efficiency and loop order
- 12:09Cuttle Explorer for code analysis
- 16:24Optimizing Offset class
- 22:09Efficient bit manipulation
- 29:31HashMaps vs ScatterMaps
- 35:56Practical optimizations
- 38:26Immediate Compose improvements
related talks
38mLow effort, High impact Android performance wins
35mThe Impact of Micro-Optimizations on Large-Scale Apps
01h 03mApplication Performance Optimisation in Practice
35mEnhancing Jetpack Compose App Performance:Tools, Strategies and Optimizations
45mPerformance in a Kotlin world
42mPerformance Paradox: Navigating the Complexities of Mobile App Optimisation
46mThe Impact of Micro-Optimizations on Large-Scale Apps
39mAndroid App Performance in a Nutshell