-
公开(公告)号：US20200274733A1
公开(公告)日：2020-08-27
申请号：US16789458
申请日：2020-02-13
Applicant: MELLANOX TECHNOLOGIES TLV LTD.
Inventor： Richard Graham , Lion Levi , Gil Bloch , Daniel Marcovitch , Noam Bloch , Yong Qin , Yaniv Blumenfeld , Eitan Zahavi
Abstract: A method in which a plurality of process are configured to hold a block of data destined for other processes, with data repacking circuitry including receiving circuitry configured to receive at least one block of data from a source process of the plurality of processes, the repacking circuitry configured to repack received data in accordance with at least one destination process of the plurality of processes, and sending circuitry configured to send the repacked data to the at least one destination process of the plurality of processes, receiving a set of data for all-to-all data exchange, the set of data being configured as a matrix, the matrix being distributed among the plurality of processes, and transposing the data by each of the plurality of processes sending matrix data from the process to the repacking circuitry, and the repacking circuitry receiving, repacking, and sending the resulting matrix data to destination processes.
-
公开(公告)号：US11876642B2
公开(公告)日：2024-01-16
申请号：US17495824
申请日：2021-10-07
Applicant: MELLANOX TECHNOLOGIES TLV LTD.
Inventor： Richard Graham , Lion Levi , Gil Bloch , Daniel Marcovitch , Noam Bloch , Yong Qin , Yaniv Blumenfeld , Eitan Zahavi
CPC classification number: H04L12/40182 , G06F12/0246 , H04B7/0456 , H04L12/44 , H04W24/10 , H04W88/06
Abstract: A method in which a plurality of process are configured to hold a block of data destined for other processes, with data repacking circuitry including receiving circuitry configured to receive at least one block of data from a source process of the plurality of processes, the repacking circuitry configured to repack received data in accordance with at least one destination process of the plurality of processes, and sending circuitry configured to send the repacked data to the at least one destination process of the plurality of processes, receiving a set of data for all-to-all data exchange, the set of data being configured as a matrix, the matrix being distributed among the plurality of processes, and transposing the data by each of the plurality of processes sending matrix data from the process to the repacking circuitry, and the repacking circuitry receiving, repacking, and sending the resulting matrix data to destination processes.
-
公开(公告)号：US20220029854A1
公开(公告)日：2022-01-27
申请号：US17495824
申请日：2021-10-07
Applicant: MELLANOX TECHNOLOGIES TLV LTD.
Inventor： Richard Graham , Lion Levi , Gil Bloch , Daniel Marcovitch , Noam Bloch , Yong Qin , Yaniv Blumenfeld , Eitan Zahavi
Abstract: A method in which a plurality of process are configured to hold a block of data destined for other processes, with data repacking circuitry including receiving circuitry configured to receive at least one block of data from a source process of the plurality of processes, the repacking circuitry configured to repack received data in accordance with at least one destination process of the plurality of processes, and sending circuitry configured to send the repacked data to the at least one destination process of the plurality of processes, receiving a set of data for all-to-all data exchange, the set of data being configured as a matrix, the matrix being distributed among the plurality of processes, and transposing the data by each of the plurality of processes sending matrix data from the process to the repacking circuitry, and the repacking circuitry receiving, repacking, and sending the resulting matrix data to destination processes.
-
公开(公告)号：US11196586B2
公开(公告)日：2021-12-07
申请号：US16789458
申请日：2020-02-13
Applicant: MELLANOX TECHNOLOGIES TLV LTD.
Inventor： Richard Graham , Lion Levi , Gil Bloch , Daniel Marcovitch , Noam Bloch , Yong Qin , Yaniv Blumenfeld , Eitan Zahavi
Abstract: A method in which a plurality of process are configured to hold a block of data destined for other processes, with data repacking circuitry including receiving circuitry configured to receive at least one block of data from a source process of the plurality of processes, the repacking circuitry configured to repack received data in accordance with at least one destination process of the plurality of processes, and sending circuitry configured to send the repacked data to the at least one destination process of the plurality of processes, receiving a set of data for all-to-all data exchange, the set of data being configured as a matrix, the matrix being distributed among the plurality of processes, and transposing the data by each of the plurality of processes sending matrix data from the process to the repacking circuitry, and the repacking circuitry receiving, repacking, and sending the resulting matrix data to destination processes.
-
-
-
Search Results
4
records
(took 0.012 seconds)
