US Patent Issued to Google on Feb. 17 for "Synthesizing zero-loss low-power approximate DNN accelerators with large-scale search" (California Inventors)

ALEXANDRIA, Va., Feb. 17 -- United States Patent no. 12,554,675, issued on Feb. 17, was assigned to Google LLC (Mountain View, Calif.).

"Synthesizing zero-loss low-power approximate DNN accelerators with large-scale search" was invented by Azalia Mirhoseini (Mountain View, Calif.), Safeen Huda (San Jose, Calif.), Martin Christoph Maas (Mountain View, Calif.), Paras Jagdish Jain (Cupertino, Calif.) and Jeffrey Adgate Dean (Palo Alto, Calif.).

According to the abstract* released by the U.S. Patent & Trademark Office: "Systems and methods are provided for designing approximate, low-power deep learning accelerator chips that have little to no accuracy loss when executing a deep learning model. A set of approximate systolic arrays may be gener...