Analytical macromodeling for high-level power estimation

Abstract

This paper presents a new macromodeling technique for high-level power estimation. Our technique is based on a parameterizable analytical model that relies exclusively on statistical information of the circuit's primary inputs. During estimation, the statistics of the required metrics are extracted from the input stream, and a power estimate is obtained by evaluating a model function that has been characterized in advance. Our model yields power estimates within seconds, because it does not rely on the statistics of the circuit's primary outputs and, consequently, does not perform any simulation during estimation. Moreover, it achieves better accuracy than previous macromodeling approaches by taking into account both spatial and temporal correlations in the input stream.In experiments with the ISCAS-85 combinational circuits, the average absolute relative error of our power macromodeling technique was at most 1.8%. The worst-case error was at most 12.8%. For a ripple-carry adder family, in comparison with power estimates that were obtained using Spice, the average absolute and worst-case errors of our model's estimates were at most 5.1% and 19.8%, respectively.In addition to power dissipation, our macromodeling technique can be used to estimate the statistics of a circuit's primary outputs with very low average errors. It is thus suitable for power estimation in core-based systems with pre-characterized blocks. Once the metrics of the primary inputs are known, the power dissipation of the entire system can be estimated by simply propagating this information through the blocks using their corresponding model functions.

A Markov Chain Sequence Generator for Power Macromodeling

“…Input signal entropy was proposed in [20] and applied to power macromodeling in [13]. Temporal correlation was introduced in [12] and applied to power macromodeling in [5]. In this paper, we choose average input signal probability p [4,15,18], average input transition density d [4,15], and input spatial correlation s [4,15] as the input parameters to generate the macromodels.…”

“…Zhanping et al introduce the concept of power sensitivity which can be used to analyze and improve the accuracy of interpolation [8,9]. Analytical power macromodeling uses mathematical expressions to calculate estimates and, therefore, avoids the space cost [5,16]. The commonly used expressions are low-order polynomial functions.…”

“…There is little analysis on the uniformity and statistics coverage of the generated sequences. In [5], a large number of sequences was generated based solely on average probability p. The generated sequences were inspected for other relevant statistics, and only sequences whose statistics were closest to the required ones were selected. Therefore, this approach is not guaranteed to cover the entire space of statistics, nor it controls the uniformity of the sequence statistics distribution.…”

Macro-models for high-level area and power estimation on FPGAs

“…An equation-based macro-modeling technique is proposed in [10]. This model uses an equation to represent the entries of an LUT and thus avoids its potentially large space requirements.…”

“…In [10], a random number generator is chosen to generate input sequence. Since there is no constraint for P in , D in and S in , some sequences may have the same P in , D in and S in and some values of P in , D in and S in may never be generated.…”