Hi canalLUZ,

Aximmetry supports modern NVIDIA GPUs, so this does not appear to be a GPU compatibility issue. Based on your description, the problem sounds more likely to be caused by real-time pipeline overload or an incorrect configuration.

Your GPU should be capable of handling more complex scenes, as you mentioned, but to confidently rule out insufficient GPU resources, you can run a few isolated tests. 
It is possible that Eye’s tracking moves the virtual camera in a way that increases the rendering load. For example, the camera may initially be facing a simple object that is easy to render, but once tracking is active, it might move toward or turn to face a more complex part of the scene. This could make rendering issues appear only after Eye is connected.

To check if this is the case, first disconnect Eye and load the scenes where the problems occurred. Use Free Camera mode to move through the entire scene while monitoring resource usage. This will help confirm whether your system can render the most demanding parts of the scene without Eye connected.
Then, test the opposite scenario, by connecting your PC to Eye and open a simple test scene. You can use the AR example scene included in the Tutorials project folder provided with your Aximmetry installation: Tutorials/AR/AR_example.xcomp. Check whether the same performance issues still occur.
These steps will help you determine whether the issue is truly related to connecting Eye.

This issue is unlikely to be caused by recent NVIDIA drivers. However, we recommend updating Aximmetry to the latest available release so you have the most recent stability fixes. 
You should also check whether Windows Power Saving mode is enabled, as it can limit the resources available to the system.

The robotic NDI audio is most likely not a separate problem, but a symptom and indicator of system overload. 
If the issue were caused by network overload, the excess data would simply not arrive in Aximmetry, and the software would warn you about missing packets. Network overload should not normally cause the kind of CPU and GPU overload you described. Even so, we strongly recommend using a wired connection between Eye and your PC to ensure the most stable connection.

If you are using Eye for video streaming and receiving the preview stream, lowering the stream resolution, frame rate and bitrate shouldn't have much effect on resource usage.
Which of the three Connection Modes in Eye are you currently using? 
Please test Eye with tracking only, without the video stream. Then enable the video stream and see if you experience any changes in performance.

Is the device you are using with Eye overheating? Overheating can cause stuttering in the app, so please ensure that the device is kept at an optimal operating temperature. Eye will display warning messages if overheating occurs.

Also, what rendering resolution and frame rate are you using in Aximmetry? Lowering these settings can help reduce the overall system load.

The CPU may also be the bottleneck. If the CPU cannot process the workload of your setup, it may overload when Eye is connected. Once the CPU is overloaded, the GPU can also be affected. When you said that system resource usage spikes beyond 100% and both CPU and GPU are affected, did you mean the Processor Load panel in Aximmetry?

Warmest regards,
Adam