Page 1 of 1

Anomaly in Force Sensor Graphs and Performance Improvement with External Monitor

Posted: 16 Feb 2025, 03:35
by zorro
I've discovered a peculiar phenomenon related to force sensor graphs. I use two figure windows to display the three forces and three torque components from the sensor at the end of a robotic arm. Due to complex external force contacts, the graph content experiences sudden changes and significant jittering when these forces occur, transforming what should be smooth graphical representations into jittery and dense visuals. During such episodes of jitter on my laptop, I noticed that simulation performance significantly degrades - characterized by a noticeable drop in frame rates and increased duration for "Main & Simulation scripts called". It was my hypothesis that this complexity in rendering adversely impacted the program's operational efficiency.

However, today while running the simulation with my laptop connected to an external monitor, I was surprised to find that the previously experienced lagging during the phase of force graph jittering had disappeared. This suggests that the earlier observed lags might be due to some sort of bug or limitation.

Here are my laptop's specifications:

CPU: i7-12700H
GPU: RTX3060
Laptop Display: 2560x1440, 165Hz
External Monitor: 2560x1440, 144Hz

Re: Anomaly in Force Sensor Graphs and Performance Improvement with External Monitor

Posted: 16 Feb 2025, 16:51
by coppelia
Hello,

The way graph items are rendered depends on the attached hardware and system configuration. When an external monitor is connected, the system might switch to a different graphics card (e.g., from an integrated GPU to a dedicated GPU), leading to performance changes. This could explain why the lag disappeared when using an external monitor. Checking GPU utilization via task manager (Windows) or activity monitor (Mac) could help confirm if a different GPU is being used in each scenario.

Also, you can specify how frequent a graph will update its content, and improve performance in that way.

Cheers

Re: Anomaly in Force Sensor Graphs and Performance Improvement with External Monitor

Posted: 18 Feb 2025, 05:03
by zorro
coppelia wrote: 16 Feb 2025, 16:51 Hello,

The way graph items are rendered depends on the attached hardware and system configuration. When an external monitor is connected, the system might switch to a different graphics card (e.g., from an integrated GPU to a dedicated GPU), leading to performance changes. This could explain why the lag disappeared when using an external monitor. Checking GPU utilization via task manager (Windows) or activity monitor (Mac) could help confirm if a different GPU is being used in each scenario.

Also, you can specify how frequent a graph will update its content, and improve performance in that way.

Cheers
You're correct. I found that my laptop was not set to use the direct GPU connection. This means that when no external monitor is connected, it resorts to using the integrated Intel graphics card. Upon connecting an external monitor, however, it utilizes the dedicated graphics card. Thank you for your help!