Android Developers Backstage

Episode 183: Baseline Profiles

Feb 25, 2022

41:56

Creator website

Highlights

AI Chapters

Episode notes

Android Gradle Plugin

03:03

Compiling Compose Code to Compiled View Code?

02:45

Adding a Baseline Profile Shares a Millisecond Per Frame

01:42

Introduction

2min

What's the Level of Compilation That Happens?

2min

What Is AOT Compilation?

2min

Is JIT Compiling Slower?

2min

The Magic of Business Profiles

2min

Aot

2min

Using Baseline Profiles to Optimize the Runtime

2min

Is Baseline Profiles a Go Fast Feature?

2min

Can Profiles Reduce Compiled Code Over Time?

2min

10.

How Cloud Profiles Can Improve Google Maps Startup Performance

2min

11.

Jetpack Compose and Baseline Profiles

2min

12.

Compiling Compose Code to Compiled View Code?

3min

13.

Is There an Opportunity to Scale to the World?

2min

14.

Android Gradle Plugin

3min

15.

Do You Offer a Way to Generate Profiles?

2min

16.

Macro Benchmarks - How Does It Work?

2min

17.

Macro Benchmark - What's the Difference?

2min

18.

How to Use the Benchmark Library to Benchmark the APK?

2min

19.

Is There a Tool I Can Run by Hand to Compile an Android App?

2min

20.

Jetpack - Is There a Baseline Profile?

2min

21.

Is Jetpack a Good Tool to Optimize Performance?

2min

22.

Adding a Baseline Profile Shares a Millisecond Per Frame

2min

In this episode we chat with Rahul and Leland from the Android toolkit team about "baseline profiles", a new mechanism which lets applications ship with profiling metadata used at application install time to make it start faster. We discuss how this is used in Jetpack Compose, cloud profiles in general, and more.

Chet, Romain, Rahul, Leland and Tor

Links:

Rahul: @tikurahul

Leland: @intelligibabble

Chet: @chethaase

Romain: @romainguy

Tor: @tornorbye