公开(公告)号：US20180247197A1

公开(公告)日：2018-08-30

申请号：US15965742

申请日：2018-04-27

申请人： Google LLC

发明人： Paul A. Tucker ,  Jeffrey Adgate Dean ,  Sanjay Ghemawat ,  Yuan Yu

IPC分类号： G06N3/08 ,  G06N5/04

CPC分类号： G06N3/08 ,  G06F9/5038 ,  G06F9/5066 ,  G06N3/0454 ,  G06N3/063 ,  G06N3/084 ,  G06N5/048

摘要： Methods, systems, and apparatus, including computer programs encoded on computer storage media, for receiving a request from a client to process a computational graph; obtaining data representing the computational graph, the computational graph comprising a plurality of nodes and directed edges, wherein each node represents a respective operation, wherein each directed edge connects a respective first node to a respective second node that represents an operation that receives, as input, an output of an operation represented by the respective first node; identifying a plurality of available devices for performing the requested operation; partitioning the computational graph into a plurality of subgraphs, each subgraph comprising one or more nodes in the computational graph; and assigning, for each subgraph, the operations represented by the one or more nodes in the subgraph to a respective available device in the plurality of available devices for operation.

公开(公告)号：US20240160948A1

公开(公告)日：2024-05-16

申请号：US18235632

申请日：2023-08-18

申请人： Google LLC

发明人： Paul A. Tucker ,  Jeffrey Adgate Dean ,  Sanjay Ghemawat ,  Yuan Yu

IPC分类号： G06N3/098 ,  G06F9/50 ,  G06N3/045 ,  G06N3/063 ,  G06N3/08 ,  G06N3/084 ,  G06N20/00

CPC分类号： G06N3/098 ,  G06F9/5038 ,  G06F9/5066 ,  G06N3/045 ,  G06N3/063 ,  G06N3/08 ,  G06N3/084 ,  G06N20/00 ,  G06N5/048

摘要： Methods, systems, and apparatus, including computer programs encoded on computer storage media, for receiving a request from a client to process a computational graph; obtaining data representing the computational graph, the computational graph comprising a plurality of nodes and directed edges, wherein each node represents a respective operation, wherein each directed edge connects a respective first node to a respective second node that represents an operation that receives, as input, an output of an operation represented by the respective first node; identifying a plurality of available devices for performing the requested operation; partitioning the computational graph into a plurality of subgraphs, each subgraph comprising one or more nodes in the computational graph; and assigning, for each subgraph, the operations represented by the one or more nodes in the subgraph to a respective available device in the plurality of available devices for operation.

公开(公告)号：US10860925B2

公开(公告)日：2020-12-08

申请号：US15337744

申请日：2016-10-28

申请人： Google LLC

发明人： Paul A. Tucker ,  Jeffrey Adgate Dean ,  Sanjay Ghemawat ,  Yuan Yu

IPC分类号： G06N3/08 ,  G06F9/50 ,  G06N3/063 ,  G06N3/04 ,  G06N5/04

摘要： Methods, systems, and apparatus, including computer programs encoded on computer storage media, for receiving a request from a client to process a computational graph; obtaining data representing the computational graph, the computational graph comprising a plurality of nodes and directed edges, wherein each node represents a respective operation, wherein each directed edge connects a respective first node to a respective second node that represents an operation that receives, as input, an output of an operation represented by the respective first node; identifying a plurality of available devices for performing the requested operation; partitioning the computational graph into a plurality of subgraphs, each subgraph comprising one or more nodes in the computational graph; and assigning, for each subgraph, the operations represented by the one or more nodes in the subgraph to a respective available device in the plurality of available devices for operation.

公开(公告)号：US20200302302A1

公开(公告)日：2020-09-24

申请号：US16898971

申请日：2020-06-11

申请人： Google LLC

发明人： Paul A. Tucker ,  Jeffrey Adgate Dean ,  Sanjay Ghemawat ,  Yuan Yu

IPC分类号： G06N3/08 ,  G06F9/50 ,  G06N3/063 ,  G06N3/04 ,  G06N5/04

摘要： Methods, systems, and apparatus, including computer programs encoded on computer storage media, for receiving a request from a client to process a computational graph; obtaining data representing the computational graph, the computational graph comprising a plurality of nodes and directed edges, wherein each node represents a respective operation, wherein each directed edge connects a respective first node to a respective second node that represents an operation that receives, as input, an output of an operation represented by the respective first node; identifying a plurality of available devices for performing the requested operation; partitioning the computational graph into a plurality of subgraphs, each subgraph comprising one or more nodes in the computational graph; and assigning, for each subgraph, the operations represented by the one or more nodes in the subgraph to a respective available device in the plurality of available devices for operation.

公开(公告)号：US11769061B2

公开(公告)日：2023-09-26

申请号：US16898971

申请日：2020-06-11

申请人： Google LLC

发明人： Paul A. Tucker ,  Jeffrey Adgate Dean ,  Sanjay Ghemawat ,  Yuan Yu

IPC分类号： G06N3/08 ,  G06N3/098 ,  G06F9/50 ,  G06N3/084 ,  G06N3/063 ,  G06N3/045 ,  G06N20/00 ,  G06N5/048

CPC分类号： G06N3/098 ,  G06F9/5038 ,  G06F9/5066 ,  G06N3/045 ,  G06N3/063 ,  G06N3/08 ,  G06N3/084 ,  G06N20/00 ,  G06N5/048

摘要： Methods, systems, and apparatus, including computer programs encoded on computer storage media, for receiving a request from a client to process a computational graph; obtaining data representing the computational graph, the computational graph comprising a plurality of nodes and directed edges, wherein each node represents a respective operation, wherein each directed edge connects a respective first node to a respective second node that represents an operation that receives, as input, an output of an operation represented by the respective first node; identifying a plurality of available devices for performing the requested operation; partitioning the computational graph into a plurality of subgraphs, each subgraph comprising one or more nodes in the computational graph; and assigning, for each subgraph, the operations represented by the one or more nodes in the subgraph to a respective available device in the plurality of available devices for operation.

公开(公告)号：US10685017B1

公开(公告)日：2020-06-16

申请号：US14885641

申请日：2015-10-16

申请人： Google LLC

发明人： Paul A. Tucker ,  Amit Singhal ,  Eric Jackson

IPC分类号： G06F16/20 ,  G06F16/242 ,  G06F16/9535

摘要： Methods and systems for efficient query rewriting and the like are described here. One such described method comprises: offline mapping frequently-seen search queries to rewritten queries that may be better for searching; offline caching the mapping in a cache memory; and upon receiving a search query from a user similar to one of the mapped search queries, obtaining a corresponding rewritten query from the mapping in the cache memory based on predetermined conditions, and issuing a search of the rewritten query to the backend data system in order to avoid having to issue a search query to the backend data system twice while the user is online.

公开(公告)号：US10534997B2

公开(公告)日：2020-01-14

申请号：US15965742

申请日：2018-04-27

申请人： Google LLC

发明人： Paul A. Tucker ,  Jeffrey Adgate Dean ,  Sanjay Ghemawat ,  Yuan Yu

IPC分类号： G06E1/00 ,  G06E3/00 ,  G06F15/18 ,  G06G7/00 ,  G06N3/08 ,  G06F9/50 ,  G06N3/063 ,  G06N3/04 ,  G06N5/04

摘要： Methods, systems, and apparatus, including computer programs encoded on computer storage media, for receiving a request from a client to process a computational graph; obtaining data representing the computational graph, the computational graph comprising a plurality of nodes and directed edges, wherein each node represents a respective operation, wherein each directed edge connects a respective first node to a respective second node that represents an operation that receives, as input, an output of an operation represented by the respective first node; identifying a plurality of available devices for performing the requested operation; partitioning the computational graph into a plurality of subgraphs, each subgraph comprising one or more nodes in the computational graph; and assigning, for each subgraph, the operations represented by the one or more nodes in the subgraph to a respective available device in the plurality of available devices for operation.