The patent badge is an abbreviated version of the USPTO patent document. The patent badge does contain a link to the full patent document.

The patent badge is an abbreviated version of the USPTO patent document. The patent badge covers the following: Patent number, Date patent was issued, Date patent was filed, Title of the patent, Applicant, Inventor, Assignee, Attorney firm, Primary examiner, Assistant examiner, CPCs, and Abstract. The patent badge does contain a link to the full patent document (in Adobe Acrobat format, aka pdf). To download or print any patent click here.

Date of Patent:
Nov. 10, 2020

Filed:

Apr. 05, 2016
Applicant:

Baidu Usa Llc, Sunnyvale, CA (US);

Inventors:

Gregory Diamos, San Jose, CA (US);

Awni Hannun, Palo Alto, CA (US);

Bryan Catanzaro, Cupertino, CA (US);

Dario Amodei, San Francisco, CA (US);

Erich Elsen, Mountain View, CA (US);

Jesse Engel, Oakland, CA (US);

Shubhabrata Sengupta, Menlo Park, CA (US);

Assignee:

Baidu USA LLC, Sunnyvale, CA (US);

Attorney:
Primary Examiner:
Int. Cl.
CPC ...
G06N 3/04 (2006.01); G06N 3/06 (2006.01); G06N 3/08 (2006.01); G06N 3/063 (2006.01);
U.S. Cl.
CPC ...
G06N 3/04 (2013.01); G06N 3/0445 (2013.01); G06N 3/063 (2013.01); G06N 3/088 (2013.01);
Abstract

Systems and methods for a multi-core optimized Recurrent Neural Network (RNN) architecture are disclosed. The various architectures affect communication and synchronization operations according to the Multi-Bulk-Synchronous-Parallel (MBSP) model for a given processor. The resulting family of network architectures, referred to as MBSP-RNNs, perform similarly to a conventional RNNs having the same number of parameters, but are substantially more efficient when mapped onto a modern general purpose processor. Due to the large gain in computational efficiency, for a fixed computational budget, MBSP-RNNs outperform RNNs at applications such as end-to-end speech recognition.


Find Patent Forward Citations

Loading…