-
公开(公告)号：EP3783484A1
公开(公告)日：2021-02-24
申请号：EP19810599.1
申请日：2019-05-22
发明人： GAO, Xiong , LI, Wei , ZHENG, Ming , LAM, Hou Fun
IPC分类号： G06F9/48
摘要： Embodiments of this application disclose a data processing method. The method in the embodiments of this application includes: generating, by a computer device, a target task, where the target task includes a buffer application task or a buffer release task, the target task is a successive task of a first task and is a preceding task of a second task, the first task and the second task are to-be-executed tasks between which there is a sequential dependency relationship, and when the target task is the buffer application task, a buffer corresponding to the buffer application task is used when the second task is executed, or when the target task is the buffer release task, a buffer corresponding to the buffer release task is used when the first task is executed; and after a preceding task of the target task is executed and before a successive task of the target task is executed, obtaining, by the computer device based on the target task, a buffer entry corresponding to the target task, where the buffer entry is an entry including a memory size of a buffer corresponding to the target task, a memory location of the buffer, and a memory address of the buffer, and executing the target task to apply for or release the buffer corresponding to the target task.
-
公开(公告)号：EP4379564A1
公开(公告)日：2024-06-05
申请号：EP21952142.4
申请日：2021-07-31
发明人： LAM, Hou Fun , LIU, Hu , WANG, Xusheng
IPC分类号： G06F15/80
CPC分类号： G06F15/80
摘要： Embodiments of this application disclose a processor and a communication method. The processor includes a primary core and a secondary core. The primary core is configured to transfer input parameters of a first function to a secondary core. The primary core is further configured to send first information to the secondary core. The first information is used by the secondary core to obtain the first function. The secondary core is configured to obtain the first function based on the first information, and run the first function based on the input parameters of the first function. The primary core is further configured to send second information to the secondary core. The second information is used by the secondary core to obtain a second function. The secondary core is further configured to obtain the second function based on the second information. The primary core is further configured to: when determining that input parameters of the second function are the same as the input parameters of the first function, control the secondary core to run the second function based on the input parameters of the first function. When the input parameters of the second function are the same as the input parameters of the first function, overheads of transferring the input parameters of the second function by the primary core to the secondary core can be reduced, and time for the secondary core to run the second function can be reduced.
-
公开(公告)号：EP4372572A2
公开(公告)日：2024-05-22
申请号：EP24151801.8
申请日：2019-05-22
发明人： GAO, Xiong , LI, Wei , ZHENG, Ming , LAM, Hou Fun
IPC分类号： G06F13/28
CPC分类号： G06F9/4881 , G06F2209/48420130101 , G06F9/5016 , G06F2209/501120130101 , G06F9/5022
摘要： Embodiments of this application disclose a data processing method. The method in the embodiments of this application includes: generating, by a computer device, a target task, where the target task includes a buffer application task or a buffer release task, the target task is a successive task of a first task and is a preceding task of a second task, the first task and the second task are to-be-executed tasks between which there is a sequential dependency relationship, and when the target task is the buffer application task, a buffer corresponding to the buffer application task is used when the second task is executed, or when the target task is the buffer release task, a buffer corresponding to the buffer release task is used when the first task is executed; and after a preceding task of the target task is executed and before a successive task of the target task is executed, obtaining, by the computer device based on the target task, a buffer entry corresponding to the target task, where the buffer entry is an entry including a memory size of a buffer corresponding to the target task, a memory location of the buffer, and a memory address of the buffer, and executing the target task to apply for or release the buffer corresponding to the target task.
-
公开(公告)号：EP4459463A2
公开(公告)日：2024-11-06
申请号：EP24176511.4
申请日：2019-01-07
发明人： LI, Wei , GAO, Xiong , LAM, Hou Fun , MA, Tao
IPC分类号： G06F9/48
摘要： A task processing method, a processing apparatus, and a computer system are provided. Implementation of the method includes: generating, by a first processing apparatus (210), a plurality of tasks, and determining task description information of the plurality of tasks, where the task description information of the plurality of tasks is used to indicate a dependency relationship between the plurality of tasks (301); sending an instruction to a second processing apparatus (220), where the instruction includes the plurality of tasks and the task description information of the plurality of tasks (302); and receiving the instruction, and processing the plurality of tasks based on the dependency relationship between the plurality of tasks (303). The method can effectively reduce a waiting delay, fully exploit a computing capability of an acceleration chip, and improve task processing efficiency.
-
公开(公告)号：EP4372572A3
公开(公告)日：2024-07-17
申请号：EP24151801.8
申请日：2019-05-22
发明人： GAO, Xiong , LI, Wei , ZHENG, Ming , LAM, Hou Fun
CPC分类号： G06F9/4881 , G06F2209/48420130101 , G06F9/5016 , G06F2209/501120130101 , G06F9/5022
摘要： Embodiments of this application disclose a data processing method. The method in the embodiments of this application includes: generating, by a computer device, a target task, where the target task includes a buffer application task or a buffer release task, the target task is a successive task of a first task and is a preceding task of a second task, the first task and the second task are to-be-executed tasks between which there is a sequential dependency relationship, and when the target task is the buffer application task, a buffer corresponding to the buffer application task is used when the second task is executed, or when the target task is the buffer release task, a buffer corresponding to the buffer release task is used when the first task is executed; and after a preceding task of the target task is executed and before a successive task of the target task is executed, obtaining, by the computer device based on the target task, a buffer entry corresponding to the target task, where the buffer entry is an entry including a memory size of a buffer corresponding to the target task, a memory location of the buffer, and a memory address of the buffer, and executing the target task to apply for or release the buffer corresponding to the target task.
-
公开(公告)号：EP4296906A1
公开(公告)日：2023-12-27
申请号：EP21933885.2
申请日：2021-03-31
发明人： ZHU, Xiangyi , LAM, Hou Fun
摘要： Embodiments of this application disclose a synchronization method and apparatus, and relate to the field of artificial intelligence, to resolve a conventional-technology problem that synchronization among AI servers is not supported. A specific solution is as follows: A first processor creates a first synchronization object for a first synchronization event. The first synchronization object includes an identifier of a first synchronization register. A value of the first synchronization register includes a first value or a second value. The first value is used to indicate that the first synchronization event does not occur, and the second value is used to indicate that the first synchronization event occurs. The first processor includes a first central processing unit CPU. The second processor determines, based on the value of the first synchronization register, whether the first synchronization event occurs. The second processor includes a first neural-network processing unit NPU.
-
公开(公告)号：EP3786793A1
公开(公告)日：2021-03-03
申请号：EP19802616.3
申请日：2019-01-07
发明人： LI, Wei , GAO, Xiong , LAM, Hou Fun , MA, Tao
IPC分类号： G06F9/46
摘要： A task processing method, a processing apparatus, and a computer system are provided. Implementation of the method includes: generating, by a first processing apparatus (210), a plurality of tasks, and determining task description information of the plurality of tasks, where the task description information of the plurality of tasks is used to indicate a dependency relationship between the plurality of tasks (301); sending an instruction to a second processing apparatus (220), where the instruction includes the plurality of tasks and the task description information of the plurality of tasks (302); and receiving the instruction, and processing the plurality of tasks based on the dependency relationship between the plurality of tasks (303). The method can effectively reduce a waiting delay, fully exploit a computing capability of an acceleration chip, and improve task processing efficiency.
-
-
-
-
-
-
搜索结果
7 条记录 (耗时 0.015 秒)
