US Patent Issued to Advanced Micro Devices on Aug. 12 for "Method and apparatus for quantization and dequantization of neural network input and output data using processing-in-memory" (California, Massachusetts, Texas Inventors)

ALEXANDRIA, Va., Aug. 12 -- United States Patent no. 12,387,767, issued on Aug. 12, was assigned to Advanced Micro Devices Inc. (Santa Clara, Calif.).

"Method and apparatus for quantization and dequantization of neural network input and output data using processing-in-memory" was invented by Ioannis Papadopoulos (Boxorough, Mass.), Vignesh Adhinarayanan (Austin, Texas), Ashwin Aji (Santa Clara, Calif.) and Jagadish B Kotra (Austin, Texas).

According to the abstract* released by the U.S. Patent & Trademark Office: "An apparatus and method for creating less computationally intensive nodes for a neural network. An integrated circuit includes a host processor and multiple memory channels, each with multiple memory array banks. Each of the mem...