Novel Algorithms for Fast Statistical Analysis of Scaled Circuits

As VLSI technology moves to the nanometer scale for transistor feature sizes, the impact of manufacturing imperfections result in large variations in the circuit performance. Traditional CAD tools are not well-equipped to handle this scenario, since they do not model this statistical nature of the circuit parameters and performances, or if they do, the existing techniques tend to be over-simplified or intractably slow.
Novel Algorithms for Fast Statistical Analysis of Scaled Circuits draws upon ideas for attacking parallel problems in other technical fields, such as computational finance, machine learning and actuarial risk, and synthesizes them with innovative attacks for the problem domain of integrated circuits. The result is a set of novel solutions to problems of efficient statistical analysis of circuits in the nanometer regime.