Skip to content

Latest commit

 

History

History
89 lines (68 loc) · 3.78 KB

benchmark_results.rst

File metadata and controls

89 lines (68 loc) · 3.78 KB

BENCHMARK

We test different versions PyQtGraph and test different Qt libraries. We only focus on the 2D line plot capabilities with OpenGL acceleration enabled.

Two threads are set up. Thread one pushes every 10 ms 100 samples into a ring buffer with a capacity of 300.000 samples. This thread has the upper priority. The other lower thread takes care of plotting the 300.000 samples in the ring buffer at a max framerate of 50 fps.

We measure CPU & GPU load and memory consumption and check whether the max of 50 fps is maintained for plotting.

CONDITIONS

  • OpenGL acceleration is enabled for all tests with the following options set:

    import pyqtgraph as pg
pg.setConfigOptions(useOpenGL=True)
pg.setConfigOptions(antialias=True)
pg.setConfigOptions(enableExperimental=True)

  • Versions

    PyOpenGL

    PyQt5

    PySide2

    PyQt6

    PySide6

    3.1.6

    5.15.2

    5.15.2

    6.4.0

    6.3.2

  • Display scaling is set to 100% in the OS.

RESULTS: Desktop Onera

GPU NVIDIA GeForce GTX 1070
CPU AMD64 Family 23 Model 113 Stepping 0 (AMD Ryzen 5 3600)
Platform Windows-10-10.0.19041-SP0
Python 3.10.6 | packaged by conda-forge | (main, Aug 22 2022, 20:29:51) [MSC v.1929 64 bit (AMD64)]
py QT_LIB <FPS> MIN MAX <RAM MB> <CPU%> <GPU%> pyqtgraph
3.10 PyQt5 50.0 50.0 50.0 167 3.3 35.7 0.11.0 + monkeypatch
3.10 PyQt5 50.0 50.0 50.0 175 3.4 35.0 0.11.1
3.10 PyQt5 50.0 50.0 50.0 183 4.0 34.8 0.12.4
3.10 PyQt5 50.0 50.0 50.0 182 3.8 35.0 0.13.1
                 
3.10 PySide2 50.0 50.0 50.0 178 3.3 35.7 0.11.0 + monkeypatch
3.10 PySide2 50.0 49.0 50.1 185 3.2 35.8 0.11.1
3.10 PySide2 50.0 50.0 50.1 194 4.2 34.3 0.12.4
3.10 PySide2 50.0 50.0 50.1 193 4.0 35.9 0.13.1
                 
3.10 PyQt6 50.0 50.0 50.0 239 4.5 33.8 0.12.4
3.10 PyQt6 50.0 50.0 50.0 184 4.0 35.6 0.13.1
                 
3.10 PySide6 50.0 49.9 50.3 196 4.5 35.8 0.12.4
3.10 PySide6 50.0 50.0 50.1 195 4.1 36.5 0.13.1

CONCLUSION

The PyQtGraph version can have an impact on drawing performance, though in the system tested here it is not that apparent. Some integrated GPUs benefit from using the older v0.11 as it can reduce memory consumption and lead to a higher sustained frame rate (tested but not yet turned into a table and shown here). The older v0.11 can only support PyQt5 and PySide2, whereas v0.12+ also supports PyQt6 and PySide6 which offer better display scaling support.

Dennis van Gils
12-10-2022

P.S. The v0.11.0 monkeypatch and details can be found here https://github.com/Dennis-van-Gils/python-dvg-pyqtgraph-monkeypatch