Profile Arm Neural Technology workloads with the Arm Streamline GUI
Introduction
Understand NX and Project Moku with Streamline
Prepare your device and workload for Streamline
Capture a profile using the Streamline GUI
Analyze NX activity and graph operations
Next Steps
Profile Arm Neural Technology workloads with the Arm Streamline GUI
Who is this for?
This Learning Path is for Android and graphics developers who want to profile their ML for Vulkan applications on the Mali Neural Accelerator using the Arm Streamline GUI.
What will you learn?
After completing this Learning Path, you'll be able to perform the following tasks:
- Identify how neural accelerator (NX), Vulkan ML workloads, and GPU shader execution relate to a Streamline capture.
- Prepare your debuggable Android workload for Streamline capture.
- Collect and open a 30-second workload capture using the Streamline GUI.
- Inspect NX activity and graph operations using the Mali Timeline and Dispatch Graph Workloads views.
Prerequisites
Before starting, you'll need the following prerequisites:
- Your own debuggable Android app, already installed on the device, with a supported ML for Vulkan workload
- An Android 11 or later device with NX-capable Mali G2 or later hardware, Mali DDK r56p0 or later, and USB debugging access
- Basic familiarity with Android applications, Vulkan, and terminal commands
- A Windows, macOS, or supported Linux host for Arm Performance Studio and Android SDK Platform Tools
- A USB data cable to connect the Android device to your host or wireless debugging enabled
Summary
This summary was drafted with an approved AI-assisted workflow and reviewed by Arm contributors before publication. Human technical review remains part of the process so the final page reflects engineering rigor, accuracy, and Arm editorial standards.
You’ll use Arm Streamline to profile an Android ML for Vulkan workload and inspect NX activity, GPU timeline events, and graph-operation costs. First, you’ll review how Vulkan graph dispatches can use NX and shader cores. Then, you’ll prepare a debuggable app and Android device and capture about 30 seconds in the Streamline GUI. Finally, you’ll analyze the capture using the Mali Timeline and Dispatch Graph Workloads views.