Profiling Statistics and Event Tracing - 2024.1 English

Vitis Model Composer User Guide (UG1483)

Document ID
UG1483
Release Date
2024-05-30
Version
2024.1 English

You can obtain profiling data when you run your AIE simulation. Analyzing this data helps you gauge the efficiency of the kernels, the stall and active times associated with each AI Engine, and pinpoint AI Engine kernels whose performance might not be optimal. This also allows you to collect data on design latency, throughput, and bandwidth. In addition to this, you can do event trace using a formatted printf statement in the code for printing debug messages. To achieve this, you should enable the option Collect profiling statistics and enable 'printf' for debugging in the Analyze tab of the Vitis Model Composer Hub block.

Important: Using this option generates a run_summary file which is written to the aiesimulator_output folder.
Important: Enabling this option might slightly increase the overall AIE simulation time.