公开(公告)号：US10776110B2

公开(公告)日：2020-09-15

申请号：US16147696

申请日：2018-09-29

Applicant: Intel Corporation

Inventor： Jonathan Pearce ,  David Sheffield ,  Srikanth Srinivasan ,  Jeffrey Cook ,  Deborah Marr ,  Abhijit Davare ,  Asit Mishra ,  Steven Burns ,  Desmond Kirkpatrick ,  Andrey Ayupov ,  Anton Alexandrovich Sorokin ,  Eriko Nurvitadhi

IPC: G06F9/30 ,  G06F9/38 ,  G06F17/16 ,  G06F7/57 ,  G06F12/0831 ,  G06F12/084

Abstract: An apparatus and method for performing efficient, adaptable tensor operations. For example, one embodiment of a processor comprises: front end circuitry to schedule a plurality of matrix operations responsive to a tensor matrix multiplication instruction; a plurality of lanes to perform parallel execution of the matrix operations, each lane comprising: first, second, and third tile registers to store blocks of a first matrix (A), second matrix (B), and third matrix (C), respectively; at least one tensor arithmetic logic unit (TALU) to multiply a block of elements of the first matrix with a block of elements of the second matrix to generate a product and to accumulate the product with a block of elements of the third matrix, wherein each lane is to multiply one or more different blocks of the first and second matrix and to accumulate the resulting one or more products with one or more different blocks of the third matrix; and broadcast circuitry to broadcast one or more invariant matrix blocks to different tile registers within a lane and/or different tile registers across different lanes.

公开(公告)号：US12197601B2

公开(公告)日：2025-01-14

申请号：US17560193

申请日：2021-12-22

Applicant: Intel Corporation

Inventor： Ren Wang ,  Sameh Gobriel ,  Somnath Paul ,  Yipeng Wang ,  Priya Autee ,  Abhirupa Layek ,  Shaman Narayana ,  Edwin Verplanke ,  Mrittika Ganguli ,  Jr-Shian Tsai ,  Anton Sorokin ,  Suvadeep Banerjee ,  Abhijit Davare ,  Desmond Kirkpatrick ,  Rajesh M. Sankaran ,  Jaykant B. Timbadiya ,  Sriram Kabisthalam Muthukumar ,  Narayan Ranganathan ,  Nalini Murari ,  Brinda Ganesh ,  Nilesh Jain

IPC: G06F15/78 ,  G06F9/50 ,  G06F21/62 ,  G06F21/72

Abstract: Examples described herein relate to offload circuitry comprising one or more compute engines that are configurable to perform a workload offloaded from a process executed by a processor based on a descriptor particular to the workload. In some examples, the offload circuitry is configurable to perform the workload, among multiple different workloads. In some examples, the multiple different workloads include one or more of: data transformation (DT) for data format conversion, Locality Sensitive Hashing (LSH) for neural network (NN), similarity search, sparse general matrix-matrix multiplication (SpGEMM) acceleration of hash based sparse matrix multiplication, data encode, data decode, or embedding lookup.

公开(公告)号：US20200104126A1

公开(公告)日：2020-04-02

申请号：US16147696

申请日：2018-09-29

Applicant: Intel Corporation

Inventor： Jonathan Pearce ,  David Sheffield ,  Srikanth Srinivasan ,  Jeffrey Cook ,  Deborah Marr ,  Abhijit Davare ,  Asit Mishra ,  Steven Burns ,  Desmond Kirkpatrick ,  Andrey Ayupov ,  Anton Alexandrovich Sorokin ,  Eriko Nurvitadhi

IPC: G06F9/30 ,  G06F9/38 ,  G06F17/16 ,  G06F12/0831 ,  G06F12/084 ,  G06F7/57

Abstract: An apparatus and method for performing efficient, adaptable tensor operations. For example, one embodiment of a processor comprises: front end circuitry to schedule a plurality of matrix operations responsive to a tensor matrix multiplication instruction; a plurality of lanes to perform parallel execution of the matrix operations, each lane comprising: first, second, and third tile registers to store blocks of a first matrix (A), second matrix (B), and third matrix (C), respectively; at least one tensor arithmetic logic unit (TALU) to multiply a block of elements of the first matrix with a block of elements of the second matrix to generate a product and to accumulate the product with a block of elements of the third matrix, wherein each lane is to multiply one or more different blocks of the first and second matrix and to accumulate the resulting one or more products with one or more different blocks of the third matrix; and broadcast circuitry to broadcast one or more invariant matrix blocks to different tile registers within a lane and/or different tile registers across different lanes.