公开(公告)号：US20220113943A1

公开(公告)日：2022-04-14

申请号：US17455100

申请日：2021-11-16

Applicant: BEIJING BAIDU NETCOM SCIENCE TECHNOLOGY CO., LTD.

Inventor： Guanglai Deng ,  Chao Tian

IPC: G06F7/544 ,  G06F7/487 ,  G06F7/485 ,  G06F5/01 ,  G06N3/063

Abstract: The disclosure provides a method for multiply-add operations for a neural network. The method includes: determining types of respective pieces of data to be calculated based on a multiply-add operation request; in a condition of the type of each piece of the data to be calculated is a type of single-precision floating point, compressing mantissa of each pieces of the data to be calculated to obtain each compressed mantissa; splitting each compressed mantissa according to a preset rule and determining high digits and low digits of the compressed mantissa; and performing a multiply-add operation on each compressed mantissa based on the high digits and low digits of the compressed mantissa.