Performance profiling in Unity - Step-by-Step Execution

Concept Flow - Performance profiling

Start Profiling

↓

Run Game Code

↓

Collect Performance Data

↓

Analyze Data

↓

Identify Bottlenecks

↓

Optimize Code

↓

Repeat Profiling

↓

End Profiling

Start profiling, run the game, collect data, analyze to find slow parts, optimize, and repeat until performance improves.

Execution Sample

Unity

Profiler.BeginSample("MyFunction");
// Code to profile
Profiler.EndSample();

This code marks the start and end of a profiling sample to measure performance of 'MyFunction'.

Execution Table

Step	Action	Profiler State	Data Collected	Result
1	Profiler.BeginSample("MyFunction") called	Profiling started for MyFunction	No data yet	Ready to measure
2	Game code inside MyFunction runs	Profiling active	Execution time accumulating	Tracking performance
3	Profiler.EndSample() called	Profiling stopped for MyFunction	Total time recorded	Sample complete
4	Analyze collected data	Profiler idle	Execution time and call count available	Identify slow parts
5	Optimize code based on data	Profiler idle	No new data yet	Code improved
6	Repeat profiling to verify	Profiling restarted	New data collected	Check improvements
7	Profiling ends	Profiler stopped	Final performance data	Performance optimized

💡 Profiling ends after collecting and analyzing performance data to optimize code.

Variable Tracker

Variable	Start	After Step 2	After Step 3	After Step 6	Final
Profiler State	Idle	Active (MyFunction)	Idle	Active (MyFunction)	Stopped
Execution Time	0 ms	Accumulating	Recorded (e.g. 15 ms)	Accumulating	Final (e.g. 12 ms)
Call Count	0	Incrementing	Recorded (e.g. 1)	Incrementing	Final (e.g. 1)

Key Moments - 3 Insights

Why do we need to call Profiler.BeginSample and Profiler.EndSample around the code?

These calls mark the start and end of the code section to measure. Without them, the profiler won't know which part to track, as shown in steps 1 and 3 of the execution_table.

What does the profiler measure during the active state?

During the active state (step 2), the profiler collects execution time and call count data for the marked code, which helps identify slow parts as seen in step 4.

Why do we repeat profiling after optimizing code?

Repeating profiling (step 6) verifies if the optimizations improved performance by comparing new data with previous results, ensuring changes had the desired effect.

Visual Quiz - 3 Questions

Look at the execution_table at step 3, what does the profiler do?

A Starts measuring performance

B Analyzes the data collected

C Stops measuring performance and records data

D Optimizes the code automatically

💡 Hint

Check the 'Profiler State' and 'Result' columns at step 3 in the execution_table.

According to variable_tracker, what is the 'Profiler State' after step 6?

A Active (MyFunction)

B Idle

C Stopped

D Not started

💡 Hint

Look at the 'Profiler State' row under 'After Step 6' in variable_tracker.

If we forget to call Profiler.EndSample(), what will happen according to the flow?

A Profiler will never start

B Profiler will keep measuring indefinitely

C Profiler will stop automatically after some time

D Profiler will throw an error immediately

💡 Hint

Refer to the concept_flow where profiling starts and ends explicitly with BeginSample and EndSample.

Concept Snapshot

Performance Profiling in Unity:
- Use Profiler.BeginSample("Name") before code to measure
- Run the code section
- Use Profiler.EndSample() after code to stop measuring
- Analyze collected data to find slow parts
- Optimize code and repeat profiling
- Helps improve game performance step-by-step

Full Transcript

Performance profiling in Unity involves marking sections of code with Profiler.BeginSample and Profiler.EndSample to measure how long they take to run. The profiler collects data like execution time and call count while the code runs. After stopping the sample, you analyze the data to find slow parts or bottlenecks. Then you optimize the code and profile again to check if performance improved. This cycle repeats until the game runs smoothly. The execution table shows each step from starting profiling, running code, stopping profiling, analyzing data, optimizing, and repeating. The variable tracker shows how profiler state and data change during these steps. Key moments clarify why marking code sections is necessary and why repeating profiling is important. The quiz tests understanding of profiler states and the importance of proper start and stop calls.