When designing DSP circuits, it is important to predict their power consumption early in the design flow in order to reduce the repetition of time consuming design phases. High-level modelling is required for fast power estimation when a design is modified at the algorithm level. This paper presents a novel high-level analytical approach to estimate logic power consumption of arithmetic components implemented in FPGAs. In particular, models of adders and multipliers are presented in detail. The proposed methodology considers input signal correlation and glitching produced inside the component. It is based on an analytical computation of the switching activity in the component which takes into account the component architecture. The complete model can estimate the power consumption for any given clock frequency, signal statistics and operands’ word-lengths. Compared to other proposed power estimation methods, the number of circuit simulations needed for characterizing the power model of the component is highly reduced. The accuracy of the model is within 10% of low-level power estimates given by the tool XPower, and it achieves better overall performance.
展开▼
