Filters

20202

Settings

Now showing1 - 2 of 2
  • Some of the metrics are blocked by your 
    consent settings
    Item type:Publication, Access status: without files ,
    Bayesian Optimization for Signal Transmission including Crosstalk in a Via Array
    (2020-09-23)
    Scharff, Katharina
    ;
    Torun, Hakki M.
    ;
    Yang, Cheng
    ;
    Swaminathan, Madhavan
    ;
    Schuster, Christian
    Signal integrity is becoming an important aspect of EMC engineering of digital systems. The signal integrity performance of printed circuit board links is influenced by a large number of design choices. Bayesian Optimization is a method that can optimize a large number of parameters in a reasonable amount of time. In this work it is shown that Bayesian Optimization can be used for the optimization of interconnect models with respect to their transmission and crosstalk performance. The first example is the optimization of the transmission of a single via. The second example is a 10 by 10 via array where the optimization includes transmission and crosstalk. The results obtained with the Bayesian Optimization are compared with the results of a Genetic Algorithm and with design guidelines derived from a physical perspective.
    Publicationtype: Conference Paper
      74
  • Some of the metrics are blocked by your 
    consent settings
    Item type:Publication, Access status: without files ,
    Comparison of collaborative versus extended artificial neural networks for PDN design
    (IEEE, 2020-05)
    Schierholz, Morten
    ;
    Yang, Cheng
    ;
    Roy, Kallol
    ;
    Swaminathan, Madhavan
    ;
    Schuster, Christian
    Currently machine learning (ML) tools are not capable to provide analysis solutions for complex printed circuit boards. It is unknown how to prepare the data and how to determine the optimal architecture of the ML process. We show that both collaborative and extended artificial neural networks (ANNs) are capable to compensate drops in accuracies for predicting target impedance violations in an extended design space. It is proven that the extended ANN has the advantage of requiring less samples during the training process compared with the collaborative approach. The necessity of either approach is highly depending on the design space and the influence of the variation on the power delivery network.
    Publicationtype: Conference Paper
      113