Probabilistic Modelling of Nanoscale Inverters

of the neural network and a series of required input-output pairs are used for training the network. A. Datta et a1 [6] presented an analysis model to estimated yield for a pipelined design, which is based on the delay distributions of individual pipestages. Bahar [7] suggested an architecture based on the Markov Abstract The device failure must be taken into account in the nano-scale design. This paper presents the probabilistic logic model to ~ M M the Probabilistic behavior of a nanoscale inverter and inverter cascade. The analysis shows that the device probability distribution highly depends on the system structures and Other performance Random Field (MRF). The MRF is a powerful approach for probability system analysis. It can be used in a network with immunity to architecture failures. parameters.