Title:A high-level model of embedded flash energy consumption

Abstract:The alignment of code in deeply embedded SoCs can have a large impact on the total energy consumption. We investigate the effect of code alignment in six SoCs and find that a large proportion of this energy (up to 15% of total SoC energy consumption) can be saved by changing the alignment of the code.
A flexible model is created to predict the energy consumption of embedded flash memory on deeply embedded SoCs, when code is executing in place. This model uses the instruction level memory accesses performed by the processor, allowing the flash energy consumption of a sequence of instructions to be calculated. We derive the model parameters for five of the six SoCs and validate it. The error is as low as 5.0% for some platforms, with a 11% average normalized RMS deviation overall.
The scope for optimizing code alignment is explored and analysis shows that over 30% of loops could be better aligned across a range of benchmarks and SoCs. This could significantly reduce energy while increasing code size by less than 4.0%. We conclude that this effect has potential for an effective optimization, saving significant energy in deeply embedded SoCs.