Patents by Inventor Saumya SUNEJA

Saumya SUNEJA 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).

  • Vector quantization decoding hardware unit for real-time dynamic decompression for parameters of neural networks
    Patent number: 11880759
    Abstract: Embodiments of an electronic device include an integrated circuit, a reconfigurable stream switch formed in the integrated circuit along with a plurality of convolution accelerators and a decompression unit coupled to the reconfigurable stream switch. The decompression unit decompresses encoded kernel data in real time during operation of convolutional neural network.
    Type: Grant
    Filed: February 22, 2023
    Date of Patent: January 23, 2024
    Assignees: STMICROELECTRONICS S.r.l., STMicroelectronics International N.V.
    Inventors: Giuseppe Desoli, Carmine Cappetta, Thomas Boesch, Surinder Pal Singh, Saumya Suneja
  • VECTOR QUANTIZATION DECODING HARDWARE UNIT FOR REAL-TIME DYNAMIC DECOMPRESSION FOR PARAMETERS OF NEURAL NETWORKS
    Publication number: 20230206032
    Abstract: Embodiments of an electronic device include an integrated circuit, a reconfigurable stream switch formed in the integrated circuit along with a plurality of convolution accelerators and a decompression unit coupled to the reconfigurable stream switch. The decompression unit decompresses encoded kernel data in real time during operation of convolutional neural network.
    Type: Application
    Filed: February 22, 2023
    Publication date: June 29, 2023
    Applicants: STMICROELECTRONICS S.r.l., STMicroelectronics International N.V.
    Inventors: Giuseppe DESOLI, Carmine CAPPETTA, Thomas BOESCH, Surinder Pal SINGH, Saumya SUNEJA
  • Vector quantization decoding hardware unit for real-time dynamic decompression for parameters of neural networks
    Patent number: 11593609
    Abstract: Embodiments of an electronic device include an integrated circuit, a reconfigurable stream switch formed in the integrated circuit along with a plurality of convolution accelerators and a decompression unit coupled to the reconfigurable stream switch. The decompression unit decompresses encoded kernel data in real time during operation of convolutional neural network.
    Type: Grant
    Filed: February 18, 2020
    Date of Patent: February 28, 2023
    Assignees: STMicroelectronics S.r.l., STMicroelectronics International N.V.
    Inventors: Giuseppe Desoli, Carmine Cappetta, Thomas Boesch, Surinder Pal Singh, Saumya Suneja
  • VECTOR QUANTIZATION DECODING HARDWARE UNIT FOR REAL-TIME DYNAMIC DECOMPRESSION FOR PARAMETERS OF NEURAL NETWORKS
    Publication number: 20210256346
    Abstract: Embodiments of an electronic device include an integrated circuit, a reconfigurable stream switch formed in the integrated circuit along with a plurality of convolution accelerators and a decompression unit coupled to the reconfigurable stream switch. The decompression unit decompresses encoded kernel data in real time during operation of convolutional neural network.
    Type: Application
    Filed: February 18, 2020
    Publication date: August 19, 2021
    Inventors: Giuseppe DESOLI, Carmine CAPPETTA, Thomas BOESCH, Surinder Pal SINGH, Saumya SUNEJA