Hi All
Running code from internal tightly coupled memory is a simple and logical way to ensure that its operation is optimally fast. ITC has zero wait states and so instructions can maintain 500/600MHz operation - without having to wait while code is fetched from external memory or suffer from non-deterministic delays when sometimes it can run from cache and sometimes not. (Not to mention lower QSPI power consumption and no more radiation from their connection signals).
However there can be complications with how to set up the FlexRAM to give the correct memory partitioning (or one-time-programmable eFUSE settings) and how best to copy the code to the ITC from external memory but keep everything secured (IP protected) and still allow fail-safe field-updates.
This video show how it takes just 7 minutes to take any SDK example (or your own application that is running sluggishly in XiP and doesn't yet have a secure loader concept ready for delivery) and allow it to run at full speed, be uploaded in the field and fully secured in QSPI/HyperFlash.
https://youtu.be/fnfLQ-nbscI
It is a sequel to the video showing how to do it with secure XiP (used when the code size is too big to allow internal RAM operation) - https://community.nxp.com/t5/i-MX-RT/How-much-time-does-it-take-to-add-in-field-updating-and-On-the/...
It should be popular with manages who would like to see a fast i.MX RT development possibility rather than the ones that typically drag out for months more than budgeted ; - )
Regards
Mark
[uTasker project developer for Kinetis and i.MX RT]
Contact me by personal message or on the uTasker web site to discuss professional training or product development requirements
