Performance of TFT Board 4 on Fusion for STM32 v8

Fully featured ARM compilers available on Windows, Linux, and macOS.
Post Reply
Author
Message
sebastian.kraemer
Posts: 6
Joined: 01 Mar 2022 15:59

Performance of TFT Board 4 on Fusion for STM32 v8

#1 Post by sebastian.kraemer » 30 Mar 2022 11:08

Hi,
I'm trying to get some rough numbers about how much time is needed to render on the connected touch screen.

HW:
  • Fusion for STM32 v8
  • STM32F407ZGT MCU card
  • TFT Board 4 Capacitive
SW:
  • mikroSDK
  • setup with Necto Designer
I measured the time it takes for vtft_process by reading a timer value before and afterwards. I get the following numbers:
  • refresh whole screen: 217ms
  • press "small" button: 5ms
  • press "big" button: 10ms
Questions:
  • Do these results agree with your experiences or am I measuring something incorrectly?
  • The time it takes seems to be directly proportional to the number of written pixels - is the limiting factor the Intel8080 interface between MCU and TFT? Can this somehow be optimized by not using the MCU to driving the communication but using e.g. the FSMC module?
  • Is the rendering code optimized for different platforms? E.g. on STM32 mcus with ChromeArt support does the vtft (or underlying) library make use of this hardware acceleration?
kind regards,
Sebastian

Post Reply

Return to “ARM AI Compilers”