Monitor App performance using Jetpack Benchmark
summary
Jetpack Benchmark (stable since 2022) measures real results: macrobenchmark for cold/warm/hot startup and frame timing (initial vs full display) and microbenchmark for hot functions like bitmap pixel reads, showing median/min/max and percentiles (50/90/95/99) and recommending 3–5 iterations (up to 10) for stability. Android Studio creates a benchmark module, adds a profilable tag and a benchmark build variant, and you can use gradle managed devices (consistent Pixel 6 profiles) to run repeatable CI-friendly measurements.
AI-written summary · about
chapters (15)
- 0:00Introduction and speaker intro
- 1:50Why app performance matters
- 3:55Jetpack Benchmark overview
- 5:03Macrobenchmark explained
- 7:10Microbenchmark explained
- 8:39Getting started in Android Studio
- 13:18Macrobenchmark metrics
- 16:51App startup types
- 19:20Writing a startup benchmark
- 24:19Running benchmark tests
- 26:04Interpreting results
- 28:55Trace section metrics
- 31:05Microbenchmark metrics
- 36:07Gradle managed devices
- 38:36Summary and key takeaways
related talks
33mAndroid Benchmarking and other stories
43mBenchmarking your application
33mAndroid Benchmarking and other stories
26mJetpack Benchmarks and Where to Find Them
10mMeasuring Jank and Startup with Macrobenchmark
20mMeasure What Matters with Jetpack Benchmark (Lightning)
46mDo you need the Jetpack Macrobenchmark library?
44mAndroid Benchmarking and other stories