How Can Unity Developers Identify GPU Pressure and Analyze Stutter Frames?
As game visuals become increasingly sophisticated, detailed models, advanced lighting, and complex effects continue to increase rendering pressure. To help developers investigate these issues, GameOptim’s GOT Online Overview mode now includes two new modules: GPU Counter and Timeline. GPU Counter provides deeper insight into GPU pressure, while Timeline offers a visual way to analyze stutter frames and thread activity.
GOT Online’s new GPU Counter and Timeline modules provide two complementary ways to analyze game performance: GPU Counter helps identify potential causes of high GPU pressure by analyzing GPU shader and bandwidth data on supported Mali devices. Timeline provides a visual representation of main thread and worker thread activity around stutter frames. Together, these modules help developers move from identifying a performance symptom to investigating the underlying activity. How Can GPU Counter Help Identify High GPU Pressure? High GPU pressure can contribute to issues such as increased device temperature and frequency reduction. Common sources of GPU pressure include Overdraw, complex shaders, and excessive GPU read/write activity . The GPU Counter module supports collecting GPU shading and bandwidth information on some Mali based devices during Overview data collection. These metrics can help developers identify scenes with high GPU pressure and investigate potential causes. What Does Cycles/Pixel Tell Developers? Cycles/Pixel represents the average number of GPU clock cycles spent processing each pixel. Shader complexity can have a significant impact on GPU Cycles. When shaders become highly complex, the GPU needs to spend more clock cycles performing shader calculations, which can increase GPU processing time and potentially contribute to stuttering. 【Image: GPU Counter – Cycles/Pixel analysis】 By examining Cycles/Pixel under GPU shading data, developers can quickly locate scenes with high GPU cycle usage. They can then investigate whether excessive shader complexity is contributing to GPU pressure and target optimization accordingly. How Can GPU Bandwidth Reveal GPU Pressure? Similar to CPU activity, GPU bandwidth is an important indicator related to chip power consumption. When the GPU continuously performs high load external read and write operations, power consumption can increase more quickly. GameOptim’s GPU bandwidth module measures the total read/write bandwidth per frame during testing. 【Image: GPU Counter – GPU bandwidth analysis】 By examining GPU bandwidth, developers can identify scenes with relatively high bandwidth usage during testing and investigate the underlying causes for further optimization. How Does Timeline Help Analyze Stutter Frames? GOT Online’s Overview mode already provides stutter analysis. Developers can use the call stack to investigate the specific causes of individual stutter frames. The new Timeline module provides a more visual representation of the timing and activity within a stutter frame. 【Image: GOT Online Timeline module】 What Can Developers See in Timeline? Timeline clearly shows activity on the main thread and worker threads , with up to five threads displayed. For each thread, developers can see not only parent functions but also the execution time of their child functions. The length of each function's time bar represents its execution time. A longer bar indicates higher execution time around that frame. How Can Timeline Help Locate Stutter Bottlenecks? Timeline's visual representation makes it easier to understand the relationship between activity across different threads during a stutter frame. For example, developers can investigate: How many different threads are active What operations each thread is executing How those operations are progressing Whether threads are waiting for other threads Whether any threads are idle How work is queued between threads This makes it easier to investigate the relationships between thread activity and locate stutter bottlenecks in greater depth and with greater precision. How Can GPU Counter and Timeline Complement Each Other? GPU Counter and Timeline focus on different aspects of performance analysis: | Module | Primary Focus | Key Information | | | | | | GPU Counter | GPU pressure | Cycles/Pixel, GPU bandwidth | | Timeline | Stutter analysis | Thread activity, function timing, execution relationships | GPU Counter helps developers investigate why GPU pressure may be high , while Timeline helps visualize what different threads are doing around a stutter frame . Together, they provide additional data for investigating GPU related pressure and frame level performance issues. Key Takeaways GPU Counter provides GPU shading and bandwidth information on supported Mali devices. Cycles/Pixel can help identify scenes with high GPU cycle usage and investigate potential shader related pressure. GPU bandwidth measures the total read/write bandwidth per frame during testing. Timeline visualizes main thread and worker thread activity around stutter frames. Timeline supports displaying up to five threads , including parent and child function timing. These visualized metrics can help developers investigate performance bottlenecks more precisely. FAQ What devices support GPU Counter? GPU Counter supports GPU shading and bandwidth data collection on some Mali based devices when collecting Overview data. What does Cycles/Pixel measure? Cycles/Pixel represents the average number of GPU clock cycles spent processing each pixel. What does the GPU bandwidth module measure? It measures the total GPU read/write bandwidth per frame during testing. How many threads can Timeline display? Timeline can display up to five threads , including the main thread and worker threads. Can Timeline show individual function timing? Yes. In addition to parent functions, Timeline shows the execution time of child functions within each thread. New users can get 2 hours of free GOT Online testing time after completing registration. https://www.gameoptim.com/?fopt=blog About GameOptim GOT Online GOT Online is a cloud based performance analysis platform with online performance reports for deeper analysis and collaboration. Based on visualized data display, providing targeted analysis suggestions and performance parameter trends, facilitating efficient quality control execution. To meet a wide range of optimization needs, GOT Online provides four major analysis modes: • Overall Performance Analysis • Mono Heap Memory Analysis • Lua Performance Analysis • GPU Performance Analysis