How does the performance of "Execute-in-Place (XIP)" from a QSPI NOR flash device compare to executing directly from internal SRAM?
For reference, I am using the RT1170_EVK development platform with the Zephyr RTOS. My application will be too large to fit into the ITCM/DTCM memories - so, I'd like to know if I should leverage a bootloader to move the application from flash to the internal SRAM before executing.
Thanks!
Solved! Go to Solution.
Hi @jtaylor24
The performance is generally the best when when running from internal ITC RAM, which is next to the core. Also performance is dependent to the application. AN12437 i.MX RT Series Performance Optimization section 3.3 Performance comparison on different memory gives an example for this topic.
All the best
Diego
Hi @jtaylor24
The performance is generally the best when when running from internal ITC RAM, which is next to the core. Also performance is dependent to the application. AN12437 i.MX RT Series Performance Optimization section 3.3 Performance comparison on different memory gives an example for this topic.
All the best
Diego