Status report

[peardox]

Extending pimoroni/pimoroni-pico#1004

Now added FPS which gives faster than I expected numbers

Frame #593 Time = 19727 us, FPS = 50.55

[Slion]

Awesome and yes I don't think it should be slower than the hardware SPI. That's the point of PIO.

[peardox]

It is noted in the PIO code that the original author made the PIO 8 bit and mentions that 32 bit would be more efficient but less robust (basically problems may come from only needing to transfer < 32 bits as far as I can see)

It will be worthwhile altering the test conditions a little so that the test only displays a single image which is updated every frame. i.e. an option to not rotate the image or do any of the calculations. Such a static frame buffer would therefore give an accurate measure of the maximum FPS / minimum Frame Time. A similar test should also be done using the current libraries.

This test is, in itself, completely pointless. The ST7789 has it's own frame memory so if you load it once it'll stay that way until you update it. The value comes from just timing how long it takes to transfer a complete frame (in the proposed test it'd just be the same image every time). The results of such a test and an SPI version would therefore show if there was any benefit taking over a PIO unless strictly required (for reading assuming this is possible)

I would have expected that PIO vs SPI to have comparable speeds had I not had such a startling increase in the frame rate. It is quite possible that they are equally performant but the only way to definitively discover if this is so is by these tests.