Abstract: A processor arranged to compress neural network activation data comprising an input module for obtaining neural network activation data. The processor also comprises a block creation module arranged to split the neural network activation data into a plurality of blocks; and a metadata generation module for generating metadata associated with at least one of the plurality of blocks. Based on the metadata generated a selection module selects a compression scheme for each of the plurality of blocks, and a compression module for applying the selected compression scheme to the corresponding block to produce compressed neural network activation data. An output module is also provided for outputting the compressed neural network activation data.

Abstract: Disclosed is a novel neural network architecture and methods for generating neural network-based models from such architecture. A first version of the neural network, that is used for training purposes, includes one or more blocks in a first format that can then be replaced with corresponding blocks in a second format for execution. An executable model can thus be provided comprising a second version of the neural network including the one or more blocks in the second format. This then allows the training to be performed in a first, e.g. expanded format, but with a second, e.g. reduced, format model then provided for execution.

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.

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.

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.

Abstract: A processor arranged to compress neural network activation data comprising an input module for obtaining neural network activation data. The processor also comprises a block creation module arranged to split the neural network activation data into a plurality of blocks; and a metadata generation module for generating metadata associated with at least one of the plurality of blocks. Based on the metadata generated a selection module selects a compression scheme for each of the plurality of blocks, and a compression module for applying the selected compression scheme to the corresponding block to produce compressed neural network activation data. An output module is also provided for outputting the compressed neural network activation data.

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.