Patents by Inventor Michael Ferdman

Michael Ferdman has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).

  • Patent number: 10726330
    Abstract: System, method, and accelerator to process a convolutional neural network. In accordance therewith, a tile structure having input data values is loaded for a convolution layer. Each tile of the tile structure corresponds to a respective feature map in a set of input feature maps. The tile structure of each iteration represents a different subset of data values in the input feature maps. Intermediate data values associated with a subset of the data values of the input feature maps in the current intermediate tile structure are reused, when the intermediate data values of a previous tile structure overlap values to be computed in the current tile structure. Intermediate non-overlapping data values that are associated with the subset of the data values in the current tile structure are computed using associated filters having weight data values. Available reused intermediate data values and computed intermediate data values are buffered as intermediate data.
    Type: Grant
    Filed: October 11, 2017
    Date of Patent: July 28, 2020
    Assignee: The Research Foundation for The State University of New York
    Inventors: Michael Ferdman, Peter Milder, Manoj Alwani
  • Publication number: 20190220734
    Abstract: System, method, and accelerator to process a convolutional neural network. In accordance therewith, a tile structure having input data values is loaded for a convolution layer. Each tile of the tile structure corresponds to a respective feature map in a set of input feature maps. The tile structure of each iteration represents a different subset of data values in the input feature maps. Intermediate data values associated with a subset of the data values of the input feature maps in the current intermediate tile structure are reused, when the intermediate data values of a previous tile structure overlap values to be computed in the current tile structure. Intermediate non-overlapping data values that are associated with the subset of the data values in the current tile structure are computed using associated filters having weight data values. Available reused intermediate data values and computed intermediate data values are buffered as intermediate data.
    Type: Application
    Filed: October 11, 2017
    Publication date: July 18, 2019
    Inventors: Michael Ferdman, Peter Milder, Manoj Alwani