The Android Performance Analyzer (APA) is available today in Open Beta! 🚀 It's our brand new system-wide profiler for Android devs, designed to help you analyze CPU, GPU, memory, and power usage to make apps and games run better and faster.

Choose your ideal setup. 🛠️ You can use APA as a lightweight, cross-platform standalone desktop app (Windows, macOS, Linux), or integrated directly into @AndroidStudio starting with Panda 4 canary builds.

Dive deep into performance. 🕵️‍♂️ Examine CPU cores, GPU performance counters across major hardware with side by side traces. Plus, navigate traces visually with a new screenshots feature to easily pinpoint where you need to focus.

Work smarter with AI. 🧠 APA supports Perfetto SQL skills for AI agents. You can ask Gemini high-level questions like "Why is my app startup slow?" to quickly pinpoint bottlenecks, or use it to instantly build custom SQL queries for complex trace analysis.

Fast and robust. ⚡ We've made massive speed improvements. Rendering a trace in APA is now typically 6x to 26x faster than our previous profiling tools, and it is significantly more stable when working with very large traces.

Connect code to trace with Vulkan debug markers. 🔍 APA supports debug annotations for render passes, letting you see your own Render Pass names directly in the tracks. Easily link profiler workloads back to your codebase origins.

Ready to squeeze every bit of performance out of your code? 🏎️💨 Try out the Android Performance Analyzer Beta today! ➡️ Standalone: ➡️ Android Studio Canary:

Feels like we already have a lot of tools that allow us to capture performance metrics, but I would love to see more insights into how to actually interpret these beyond the most contrived examples like a Thread.sleep.

Releasing the Android Performance Analyzer in open beta is excellent news for the developer community. Teams can now analyze system resources more effectively and deliver higher quality experiences.

But does it help me persuade mid devs to fix their memory leaks… lol