公开(公告)号：US20080071943A1

公开(公告)日：2008-03-20

申请号：US11785555

申请日：2007-04-18

Applicant: Joon-ho Song ,  Si-hwa Lee ,  Do-hyung Kim

Inventor： Joon-ho Song ,  Si-hwa Lee ,  Do-hyung Kim

IPC: G06F13/28

CPC classification number: G06F13/28

Abstract: A method, medium and system for setting a transfer unit in a data processing system. The method comprises setting a transfer unit in a data processing system which repeatedly performs a process of transmitting data stored in a first memory to a second memory in a predetermined transfer unit, processing the transmitted data stored in the second memory, and transmitting the processed data to the first memory. The method includes computing overhead on the data processing system according to the size of each of a plurality of data units available as the transfer unit; and setting a data unit, which corresponds to a minimum overhead from among the computed overheads, as the transfer unit. Accordingly, it is possible to set an optimum transfer unit according to an environment of a data processing system in order to improve the performance of the data processing system.

Abstract translation: 一种在数据处理系统中设置传送单元的方法，介质和系统。 该方法包括在数据处理系统中设置传送单元，该数据处理系统重复执行将存储在第一存储器中的数据传送到预定传送单元中的第二存储器的处理，处理存储在第二存储器中的发送数据，以及发送处理后的数据 到第一个记忆。 该方法包括根据可用作传送单元的多个数据单元中的每一个的大小来计算数据处理系统的开销; 并且将从计算出的开销中的最小开销对应的数据单元设置为传送单元。 因此，为了提高数据处理系统的性能，可以根据数据处理系统的环境来设定最佳的传送单元。

公开(公告)号：US07805548B2

公开(公告)日：2010-09-28

申请号：US11785555

申请日：2007-04-18

Applicant: Joon-ho Song ,  Si-hwa Lee ,  Do-hyung Kim

Inventor： Joon-ho Song ,  Si-hwa Lee ,  Do-hyung Kim

IPC: G06F13/00 ,  G06F3/00 ,  G06F13/28

CPC classification number: G06F13/28

Abstract: A method, medium and system for setting a transfer unit in a data processing system. The method comprises setting a transfer unit in a data processing system which repeatedly performs a process of transmitting data stored in a first memory to a second memory in a predetermined transfer unit, processing the transmitted data stored in the second memory, and transmitting the processed data to the first memory. The method includes computing overhead on the data processing system according to the size of each of a plurality of data units available as the transfer unit; and setting a data unit, which corresponds to a minimum overhead from among the computed overheads, as the transfer unit. Accordingly, it is possible to set an optimum transfer unit according to an environment of a data processing system in order to improve the performance of the data processing system.

Abstract translation: 一种在数据处理系统中设置传送单元的方法，介质和系统。 该方法包括在数据处理系统中设置传送单元，该数据处理系统重复执行将存储在第一存储器中的数据传送到预定传送单元中的第二存储器的处理，处理存储在第二存储器中的发送数据，以及发送处理后的数据 到第一个记忆。 该方法包括根据可用作传送单元的多个数据单元中的每一个的大小来计算数据处理系统的开销; 并且将从计算出的开销中的最小开销对应的数据单元设置为传送单元。 因此，为了提高数据处理系统的性能，可以根据数据处理系统的环境来设定最佳的传送单元。

公开(公告)号：US20090043987A1

公开(公告)日：2009-02-12

申请号：US11984992

申请日：2007-11-26

Applicant: Joon-ho Song ,  Si-hwa Lee

Inventor： Joon-ho Song ,  Si-hwa Lee

IPC: G06F9/30 ,  G06F15/76

CPC classification number: G06F9/544 ,  G06F9/3857 ,  H04N19/423 ,  H04N19/44 ,  H04N19/61

Abstract: Provided is an operation distribution method and system using a buffer. The operation distribution system includes a buffer, a first operation device performing a first operation and storing a result of the first operation performed by the first operation device in the buffer, and a second operation device performing a second operation using the result of the first operation stored in the buffer, thereby reducing the time required to perform operations.

Abstract translation: 提供了一种使用缓冲器的操作分配方法和系统。 操作分配系统包括缓冲器，执行第一操作并将第一操作装置执行的第一操作的结果存储在缓冲器中的第一操作装置，以及使用第一操作的结果执行第二操作的第二操作装置 存储在缓冲器中，从而减少执行操作所需的时间。

公开(公告)号：US08284836B2

公开(公告)日：2012-10-09

申请号：US12081863

申请日：2008-04-22

Applicant: Dmitri Birinov ,  Joon-ho Song ,  Doo-hyun Kim

Inventor： Dmitri Birinov ,  Joon-ho Song ,  Doo-hyun Kim

IPC: H04B1/66 ,  H04N7/12 ,  H04N11/02 ,  H04N11/04

CPC classification number: H04N19/436 ,  H04N19/43 ,  H04N19/433 ,  H04N19/61

Abstract: Provided is a motion compensation method and apparatus. The motion compensation method includes performing register setting for motion compensation of an mth macroblock of a current image, performing prediction for the mth macroblock simultaneously with performing register setting for motion compensation of an (m+1)th macroblock, and performing reconstruction for the mth macroblock based on a prediction result simultaneously with performing prediction for the (m+1)th macroblock. By parallely processing motion compensation on macroblocks, the amount of time required for motion compensation of the macroblocks can be reduced.

Abstract translation: 提供了一种运动补偿方法和装置。 运动补偿方法包括对当前图像的第m个宏块执行运动补偿的寄存器设置，同时执行第m个宏块的预测，同时执行第（m + 1）个宏块的运动补偿的寄存器设置，并对第m个宏块执行重建 宏块基于预测结果同时执行第（m + 1）个宏块的预测。 通过平行处理宏块上的运动补偿，可以减少宏块的运动补偿所需的时间量。

公开(公告)号：US20090175345A1

公开(公告)日：2009-07-09

申请号：US12081863

申请日：2008-04-22

Applicant: Dmitri Birinov ,  Joon-ho Song ,  Doo-hyun Kim

Inventor： Dmitri Birinov ,  Joon-ho Song ,  Doo-hyun Kim

IPC: H04N7/26

CPC classification number: H04N19/436 ,  H04N19/43 ,  H04N19/433 ,  H04N19/61

Abstract: Provided is a motion compensation method and apparatus. The motion compensation method includes performing register setting for motion compensation of an mth macroblock of a current image, performing prediction for the mth macroblock simultaneously with performing register setting for motion compensation of an (m+1)th macroblock, and performing reconstruction for the mth macroblock based on a prediction result simultaneously with performing prediction for the (m+1)th macroblock. By parallely processing motion compensation on macroblocks, the amount of time required for motion compensation of the macroblocks can be reduced.

Abstract translation: 提供了一种运动补偿方法和装置。 运动补偿方法包括对当前图像的第m个宏块执行运动补偿的寄存器设置，同时执行第m个宏块的预测，同时执行第（m + 1）个宏块的运动补偿的寄存器设置，并对第m个宏块执行重建 宏块基于预测结果同时执行第（m + 1）个宏块的预测。 通过平行处理宏块上的运动补偿，可以减少宏块的运动补偿所需的时间量。