SPRADD8 November 2024 F29H850TU , F29H859TU-Q1
Flash execution efficiency is important because not all code can run from zero wait-state memory. On F29H85x at 200MHz, three wait-states are needed for flash access. Further, pre-fetch and block cache mechanisms are available to mitigate the effect of the wait states, and they are enabled. Flash Efficiency Benchmarks shows flash efficiency results in percentages (%) for some benchmarks comparing F29H85x and F2837x (also three wait-states at 200MHz). For many benchmarks, running from flash is akin to running from zero wait-state memory.
Benchmark | F2837x (%) | F29H85x (%) |
---|---|---|
CFFT | 81 | 93 |
FIR | 91 | 86 |
IIR (loop) | 82 | 99 |
Signal-chain (ACI) | 95 | 97 |
Binary LUT search | 82 | 97 |