Data driven signal processing: an approach for energy efficient computing

摘要: The computational switching activity of digital CMOS circuits can be dynamically minimized by designing algorithms that exploit signal statistics. This results in processors have time-varying power requirements and perform computation on demand. An approach is presented to minimize the energy dissipation per data sample variable-load DSP systems adaptively minimizing supply voltage for each using a variable speed processor. In general, buffering filtering, spread over multiple samples averaging workload lowering further. It also shown four levels quantization combined with dithering sufficient closely emulate arbitrary levels.