US Patent Issued to Nvidia on Aug. 12 for "Efficient neural network accelerator dataflows" (California, Massachusetts, Nevada, Texas Inventors)

ALEXANDRIA, Va., Aug. 12 -- United States Patent no. 12,387,089, issued on Aug. 12, was assigned to Nvidia Corp. (Santa Clara, Calif.).

"Efficient neural network accelerator dataflows" was invented by Yakun Shao (Santa Clara, Calif.), Rangharajan Venkatesan (Sunnyvale, Calif.), Miaorong Wang (Cambridge, Mass.), Daniel Smith (Los Gatos, Calif.), William James Dally (Incline Village, Nev.), Joel Emer (Acton, Mass.), Stephen W. Keckler (Austin, Texas) and Brucek Khailany (Austin, Texas).

According to the abstract* released by the U.S. Patent & Trademark Office: "A distributed deep neural net (DNN) utilizing a distributed, tile-based architecture includes multiple chips, each with a central processing element, a global memory buffer, and a pl...