公开(公告)号：US20210192337A1

公开(公告)日：2021-06-24

申请号：US16724849

申请日：2019-12-23

Applicant: Arm Limited

Inventor： Danny Daysang Loh ,  Lingchuan Meng ,  Naveen Suda ,  Eric Kunze ,  Ahmet Fatih Inci

IPC: G06N3/08 ,  G06N3/04 ,  G06N3/063

Abstract: The present disclosure advantageously provides a heterogenous system, and a method for generating an artificial neural network (ANN) for a heterogenous system. The heterogenous system includes a plurality of processing units coupled to a memory configured to store an input volume. The plurality of processing units includes first and second processing units. The first processing unit includes a first processor and is configured to execute a first ANN, and the second processing unit includes a second processor and is configured to execute a second ANN. The first and second ANNs respectively include an input layer, at least one processor-optimized hidden layer and an output layer. The second ANN hidden layers are different than the first ANN hidden layers.

公开(公告)号：US12061967B2

公开(公告)日：2024-08-13

申请号：US17132750

申请日：2020-12-23

Applicant: Arm Limited

Inventor： John Wakefield Brothers, III ,  Kartikeya Bhardwaj ,  Alexander Eugene Chalfin ,  Danny Daysang Loh

IPC: G06N3/0464 ,  G06N3/04 ,  G06N3/08

CPC classification number: G06N3/04 ,  G06N3/0464 ,  G06N3/08

Abstract: A method of processing input data for a given layer of a neural network using a data processing system comprising compute resources for performing convolutional computations is described. The input data comprises a given set of input feature maps, IFMs, and a given set of filters. The method comprises generating a set of part-IFMs including pluralities of part-IFMs which correspond to respective IFMs, of the given set of IFMs. The method further includes grouping part-IFMs in the set of part-IFMs into a set of selections of part-IFMs. The method further includes convolving, by respective compute resources of the data processing system, the set of selections with the given set of filters to compute a set of part-output feature maps. A data processing system for processing input data for a given layer of a neural network is also described.

公开(公告)号：US11620516B2

公开(公告)日：2023-04-04

申请号：US16724849

申请日：2019-12-23

Applicant: Arm Limited

Inventor： Danny Daysang Loh ,  Lingchuan Meng ,  Naveen Suda ,  Eric Kunze ,  Ahmet Fatih Inci

IPC: G06N3/08 ,  G06N20/10 ,  G06N3/063 ,  G06N3/04 ,  G06N20/00 ,  G06N20/20

Abstract: The present disclosure advantageously provides a heterogenous system, and a method for generating an artificial neural network (ANN) for a heterogenous system. The heterogenous system includes a plurality of processing units coupled to a memory configured to store an input volume. The plurality of processing units includes first and second processing units. The first processing unit includes a first processor and is configured to execute a first ANN, and the second processing unit includes a second processor and is configured to execute a second ANN. The first and second ANNs respectively include an input layer, at least one processor-optimized hidden layer and an output layer. The second ANN hidden layers are different than the first ANN hidden layers.

公开(公告)号：US11449729B2

公开(公告)日：2022-09-20

申请号：US16676757

申请日：2019-11-07

Applicant: Arm Limited

Inventor： Lingchuan Meng ,  Danny Daysang Loh ,  Ian Rudolf Bratt ,  Alexander Eugene Chalfin ,  Tianmu Li

IPC: G06N3/04 ,  G06N3/06 ,  G06N3/08 ,  G06N3/10 ,  G06F17/15 ,  G06F17/16 ,  G06F17/18 ,  G06F30/18 ,  G06F30/20 ,  G06F30/27 ,  G06F30/33 ,  G06F30/367

Abstract: The present disclosure advantageously provides a system and a method for convolving data in a quantized convolutional neural network (CNN). The method includes selecting a set of complex interpolation points, generating a set of complex transform matrices based, at least in part, on the set of complex interpolation points, receiving an input volume from a preceding layer of the quantized CNN, performing a complex Winograd convolution on the input volume and at least one filter, using the set of complex transform matrices, to generate an output volume, and sending the output volume to a subsequent layer of the quantized CNN.

公开(公告)号：US20230376745A1

公开(公告)日：2023-11-23

申请号：US17751089

申请日：2022-05-23

Applicant: Arm Limited

Inventor： Kartikeya Bhardwaj ,  Guihong Li ,  Naveen Suda ,  Milos Milosavljevic ,  Danny Daysang Loh

IPC: G06N3/08 ,  G06N3/04

CPC classification number: G06N3/08 ,  G06N3/04

Abstract: A mechanism to control the stability and performance of weight-sharing methods for designing neural networks is provided. Network weights and architecture parameters of a super-net, including multiple sub-networks, are adjusted to reduce a loss determined, at least in part, from a sum, over layers of the sub-network, of measures of smoothness based on network weights in the layers. A sub-network of the super-net is selected dependent upon the adjusted architectural parameters.

公开(公告)号：US20200151541A1

公开(公告)日：2020-05-14

申请号：US16676757

申请日：2019-11-07

Applicant: Arm Limited

Inventor： Lingchuan Meng ,  Danny Daysang Loh ,  Ian Rudolf Bratt ,  Alexander Eugene Chalfin ,  Tianmu Li

IPC: G06N3/04 ,  G06F17/16 ,  G06F17/15

Abstract: The present disclosure advantageously provides a system and a method for convolving data in a quantized convolutional neural network (CNN). The method includes selecting a set of complex interpolation points, generating a set of complex transform matrices based, at least in part, on the set of complex interpolation points, receiving an input volume from a preceding layer of the quantized CNN, performing a complex Winograd convolution on the input volume and at least one filter, using the set of complex transform matrices, to generate an output volume, and sending the output volume to a subsequent layer of the quantized CNN.