2026-01-08 10:19 AM
2026-05-28 6:24 AM
We need real-time sensor fusion (data gets streamed in over 5 UARTs and 3 SPI buses) and multiple CNN we would want to switch between to process camera data or sensor data (as a temporal neural network), plus an RGB screen buffer (1024x1024) that needs to update at least 4 times per second. Would splitting the memory into two regions (one for sensor fusion/CNN and one for the screen buffer) be the best approach? Do you think this is feasible?
2026-05-28 6:38 AM
Hi
I can't reply, let you make calculation of BW versus use case
Alternatively you may use 2 memory on 2 different memory controller
Regards
Alex
We’re moving the ST Community to a new platform to give you a better and more reliable community experience.