Invention Grant
- Patent Title: Increased precision neural processing element
-
Application No.: US18174021Application Date: 2023-02-24
-
Publication No.: US12154027B2Publication Date: 2024-11-26
- Inventor: Amol A Ambardekar , Boris Bobrov , Kent D. Cedola , Chad Balling Mcbride , George Petre , Larry Marvin Wall
- Applicant: MICROSOFT TECHNOLOGY LICENSING, LLC
- Applicant Address: US WA Redmond
- Assignee: MICROSOFT TECHNOLOGY LICENSING, LLC
- Current Assignee: MICROSOFT TECHNOLOGY LICENSING, LLC
- Current Assignee Address: US WA Redmond
- Agency: Newport IP, LLC
- Agent Jacob P. Rohwer
- Main IPC: G06N3/063
- IPC: G06N3/063 ; G06F7/499 ; G06F7/523 ; G06F7/57 ; G06F9/30

Abstract:
Neural processing elements are configured with a hardware AND gate configured to perform a logical AND operation between a sign extend signal and a most significant bit (“MSB”) of an operand. The state of the sign extend signal can be based upon a type of a layer of a deep neural network (“DNN”) that generate the operand. If the sign extend signal is logical FALSE, no sign extension is performed. If the sign extend signal is logical TRUE, a concatenator concatenates the output of the hardware AND gate and the operand, thereby extending the operand from an N-bit unsigned binary value to an N+1 bit signed binary value. The neural processing element can also include another hardware AND gate and another concatenator for processing another operand similarly. The outputs of the concatenators for both operands are provided to a hardware binary multiplier.
Public/Granted literature
- US20230196086A1 INCREASED PRECISION NEURAL PROCESSING ELEMENT Public/Granted day:2023-06-22
Information query