Is SPI the fastest way to get data out of chip?
Posted: Tue Jul 25, 2023 5:41 pm
I'm working on a project that needs to transfer data at a moderately high and regular rate out of an ESP32S3 and into an FPGA. Just to give you a rough idea, I have 36k chunks that I need transferred AT LEAST at about 1ms per chunk. If I do the math correctly (and I'm able to hit 80MHz transfer to my FPGA), this should be possible (80e6 / (36*1024) = ~2170 or 460us). I'm guessing the 80MHz will be too fast for my FPGA, though AND regardless, the more data I can get through, the more functional my project will be.
So I have a couple of questions:
- Am I correct in assuming that SPI is the fastest path out of the ESP32S3?
- Can I get the transfers to happen in a non-blocking fashion?
- If I were to build my own board, will I be able to use one of the Quad SPI channels or are these by requirement taken up by Flash & RAM?
Thanks!
So I have a couple of questions:
- Am I correct in assuming that SPI is the fastest path out of the ESP32S3?
- Can I get the transfers to happen in a non-blocking fashion?
- If I were to build my own board, will I be able to use one of the Quad SPI channels or are these by requirement taken up by Flash & RAM?
Thanks!