McPOWER: a Monte Carlo approach to power estimation
Richard Burch, Farid N. Najm, Ping Yang, Timothy N. Trick · 1992
Excessive power dissipation in integrated circuits causes overheating and can lead to soft errors and/or permanent damage. The severity of the problem increases in proportion to the level of integration, so that power estimation tools are badly needed for present-day technology. Traditional simulation-based approaches simulate the circuit using test/functional input pattern sets. This is expensive and does not guarantee ameaningful power value. Other recent approaches have used probabilistic techniques in order to cover a large set of inputs patterns. However, they trade-off accuracy for speed in ways that are not always acceptable. In this paper, we investigate an alternative technique that combines the accuracy of simulation-based techniques with the speed of the probabilistic techniques. The resulting method is statistical in nature; it consists of applying randomly-generated input patterns to the circuit and monitoring, with a simulator, the resulting power value. This is continued until a value of power is obtained with a desired accuracy, at a specified con dence level. We present the algorithm and experimental results, and discuss the superiority of this new approach.