公开(公告)号：EP4379648A2

公开(公告)日：2024-06-05

申请号：EP24171188.6

申请日：2018-03-08

申请人： INTEL Corporation

发明人： Nurvitadhi, Eriko ,  Vembu, Balaji ,  Lin, Tsung-Han ,  Sinha, Kamal ,  Barik, Rajkishore ,  Galoppo von Borries, Nicolas

IPC分类号： G06T1/20

CPC分类号： G06F12/0811 ,  G06F12/0815 ,  G06F12/0831 ,  G06F12/0888 ,  G06F2212/102420130101 ,  G06F2212/30220130101 ,  G06F2212/62120130101 ,  G06F2212/40120130101 ,  G06T1/20 ,  G06N3/063 ,  G06N3/084 ,  Y02D10/00 ,  G06N3/044 ,  G06N3/045

摘要： A method and a data processing system for performing operations of a machine learning framework to facilitate a sparse matrix multiplication for convolutional layers are disclosed. Therefore, one or more primitives for enabling a compute kernel to specify one or more operations to be performed by the compute kernel of a graphics processor on an output of a sparse matrix multiplication operation are provided, the one or more operations including applying an activation function and the compute kernel being associated with the convolutional layers. An execution of the compute kernel to execute the sparse matrix multiplication operation on matrices relating to the convolutional layers and an execution of the compute kernel to apply the activation function to the output of the sparse matrix multiplication operation based on the one or more primitives are triggered.

公开(公告)号：EP4379648A3

公开(公告)日：2024-09-11

申请号：EP24171188.6

申请日：2018-03-08

申请人： INTEL Corporation

发明人： Nurvitadhi, Eriko ,  Vembu, Balaji ,  Lin, Tsung-Han ,  Sinha, Kamal ,  Barik, Rajkishore ,  Galoppo von Borries, Nicolas

IPC分类号： G06N3/0442 ,  G06N3/0464 ,  G06N3/048 ,  G06N3/063 ,  G06N3/084 ,  G06N3/09 ,  G06T1/20

CPC分类号： G06F12/0811 ,  G06F12/0815 ,  G06F12/0831 ,  G06F12/0888 ,  G06F2212/102420130101 ,  G06F2212/30220130101 ,  G06F2212/62120130101 ,  G06F2212/40120130101 ,  G06T1/20 ,  G06N3/063 ,  G06N3/084 ,  Y02D10/00 ,  G06N3/0464 ,  G06N3/048 ,  G06N3/09 ,  G06N3/0442

摘要： A method and a data processing system for performing operations of a machine learning framework to facilitate a sparse matrix multiplication for convolutional layers are disclosed. Therefore, one or more primitives for enabling a compute kernel to specify one or more operations to be performed by the compute kernel of a graphics processor on an output of a sparse matrix multiplication operation are provided, the one or more operations including applying an activation function and the compute kernel being associated with the convolutional layers. An execution of the compute kernel to execute the sparse matrix multiplication operation on matrices relating to the convolutional layers and an execution of the compute kernel to apply the activation function to the output of the sparse matrix multiplication operation based on the one or more primitives are triggered.

公开(公告)号：EP3977297B1

公开(公告)日：2024-04-17

申请号：EP20729430.7

申请日：2020-05-08

IPC分类号： G06F12/0831 ,  G06F12/0837

CPC分类号： G06F12/0837 ,  G06F2212/101620130101 ,  G06F2212/62120130101 ,  G06F12/0835

公开(公告)号：EP4191419B1

公开(公告)日：2024-09-11

申请号：EP22193823.6

申请日：2022-09-05

IPC分类号： G06F12/0815

CPC分类号： G06F12/0815 ,  G06F2212/62120130101

公开(公告)号：EP3398071B1

公开(公告)日：2024-05-08

申请号：EP16882624.6

申请日：2016-12-29

IPC分类号： G06F12/02 ,  G06F12/084 ,  G06F12/0813 ,  G06F12/0831

CPC分类号： G06F12/0833 ,  G06F12/0813 ,  G06F2212/62120130101 ,  G06F2212/31420130101

公开(公告)号：EP4109281B1

公开(公告)日：2024-05-01

申请号：EP22165672.1

申请日：2022-03-30

IPC分类号： G06F12/0831 ,  G06F12/0813 ,  G06F13/40 ,  G06F13/42 ,  G06T1/60

CPC分类号： G06F13/4221 ,  G06T1/60 ,  Y02D10/00 ,  G06F12/0831 ,  G06F2212/50720130101 ,  G06F2212/102420130101 ,  G06F2212/62120130101 ,  G06F12/0813 ,  G06F2213/002620130101 ,  G06F13/4022 ,  G06F13/4282