Thank you for the additional pointers, SiliconWizard and ttt! I am properly intimidated by the i.MX RT.
It's a complex beast plus a new (to me) programming environment.
Just received a Teensy 4.1 board to play with. I will probably start with some performance tests using inline assembly code within the Arduino environment. (I intend to "abuse" the processor for an emulator/accelerator for legacy CPUs, on a small DIP-40 board that replaces the CPU. Need to check the response time to the host's clock edges first -- with all the caching, pipelining and branch prediction going on in the M7, this might not be as quick as I hope?) In parallel I will try and work my way into bare metal programming. Thank you for the starting point there, ttt!
I am sure I will be back on this forum with more questions...