Abstract: Methods, systems, and computer storage media for implementing neural networks in fixed point arithmetic computing systems. In one aspect, a method includes the actions of receiving a request to process a neural network using a processing system that performs neural network computations using fixed point arithmetic; for each node of each layer of the neural network, determining a respective scaling value for the node from the respective set of floating point weight values for the node; and converting each floating point weight value of the node into a corresponding fixed point weight value using the respective scaling value for the node to generate a set of fixed point weight values for the node; and providing the sets of fixed point floating point weight values for the nodes to the processing system for use in processing inputs using the neural network.

Abstract: Methods for receiving a request to process, on a hardware circuit, a neural network comprising a first convolutional neural network layer having a stride greater than one, and in response, generating instructions that cause the hardware circuit to, during processing of an input tensor, generate a layer output tensor equivalent to an output of the first convolutional neural network layer by processing the input tensor using a second convolutional neural network layer having a stride equal to one but that is otherwise equivalent to the first convolutional neural network layer to generate a first tensor, zeroing out elements of the first tensor that would not have been generated if the second convolutional neural network layer had the stride of the first convolutional neural network layer to generate a second tensor, and performing max pooling on the second tensor to generate the layer output tensor.

Abstract: Methods and systems for receiving a request to implement a neural network comprising an average pooling layer on a hardware circuit, and in response, generating instructions that when executed by the hardware circuit, cause the hardware circuit to, during processing of a network input by the neural network, generate a layer output tensor that is equivalent to an output of the average pooling neural network layer by performing a convolution of an input tensor to the average pooling neural network layer and a kernel with a size equal to a window of the average pooling neural network layer and composed of elements that are each an identity matrix to generate a first tensor, and performing operations to cause each element of the first tensor to be divided by a number of elements in the window of the average pooling neural network layer to generate an initial output tensor.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for depth concatenation using a matrix computation unit. One of the methods includes: receiving a request to process network inputs to a neural network using an integrated circuit, the neural network comprising a depth concatenation neural network layer; and generating instructions that, when executed by the integrated circuit, cause the integrated circuit to perform operations comprising: for each spatial location in a first input tensor to the depth concatenation layer and a second input tensor to the depth concatenation layer: multiplying, using the matrix computation unit, a second depth vector for the spatial location by a shift weight matrix for the depth concatenation layer to generate a shifted second depth vector; and adding the shifted second depth vector and a first input depth vector for the spatial location to generate a concatenated depth vector.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for depth concatenation using a matrix computation unit. One of the methods includes: receiving a request to process network inputs to a neural network using an integrated circuit, the neural network comprising a depth concatenation neural network layer; and generating instructions that, when executed by the integrated circuit, cause the integrated circuit to perform operations comprising: for each spatial location in a first input tensor to the depth concatenation layer and a second input tensor to the depth concatenation layer: multiplying, using the matrix computation unit, a second depth vector for the spatial location by a shift weight matrix for the depth concatenation layer to generate a shifted second depth vector; and adding the shifted second depth vector and a first input depth vector for the spatial location to generate a concatenated depth vector.

Abstract: Methods for receiving a request to process, on a hardware circuit, a neural network comprising a first convolutional neural network layer having a stride greater than one, and in response, generating instructions that cause the hardware circuit to, during processing of an input tensor, generate a layer output tensor equivalent to an output of the first convolutional neural network layer by processing the input tensor using a second convolutional neural network layer having a stride equal to one but that is otherwise equivalent to the first convolutional neural network layer to generate a first tensor, zeroing out elements of the first tensor that would not have been generated if the second convolutional neural network layer had the stride of the first convolutional neural network layer to generate a second tensor, and performing max pooling on the second tensor to generate the layer output tensor.

Abstract: Methods for receiving a request to process, on a hardware circuit, a neural network comprising a first convolutional neural network layer having a stride greater than one, and in response, generating instructions that cause the hardware circuit to, during processing of an input tensor, generate a layer output tensor equivalent to an output of the first convolutional neural network layer by processing the input tensor using a second convolutional neural network layer having a stride equal to one but that is otherwise equivalent to the first convolutional neural network layer to generate a first tensor, zeroing out elements of the first tensor that would not have been generated if the second convolutional neural network layer had the stride of the first convolutional neural network layer to generate a second tensor, and performing max pooling on the second tensor to generate the layer output tensor.

Abstract: Methods, systems, and computer storage media for implementing neural networks in fixed point arithmetic computing systems. In one aspect, a method includes the actions of receiving a request to process a neural network using a processing system that performs neural network computations using fixed point arithmetic; for each node of each layer of the neural network, determining a respective scaling value for the node from the respective set of floating point weight values for the node; and converting each floating point weight value of the node into a corresponding fixed point weight value using the respective scaling value for the node to generate a set of fixed point weight values for the node; and providing the sets of fixed point floating point weight values for the nodes to the processing system for use in processing inputs using the neural network.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for depth concatenation using a matrix computation unit. One of the methods includes: receiving a request to process network inputs to a neural network using an integrated circuit, the neural network comprising a depth concatenation neural network layer; and generating instructions that, when executed by the integrated circuit, cause the integrated circuit to perform operations comprising: for each spatial location in a first input tensor to the depth concatenation layer and a second input tensor to the depth concatenation layer: multiplying, using the matrix computation unit, a second depth vector for the spatial location by a shift weight matrix for the depth concatenation layer to generate a shifted second depth vector; and adding the shifted second depth vector and a first input depth vector for the spatial location to generate a concatenated depth vector.

Abstract: Methods and systems for receiving a request to implement a neural network comprising an average pooling layer on a hardware circuit, and in response, generating instructions that when executed by the hardware circuit, cause the hardware circuit to, during processing of a network input by the neural network, generate a layer output tensor that is equivalent to an output of the average pooling neural network layer by performing a convolution of an input tensor to the average pooling neural network layer and a kernel with a size equal to a window of the average pooling neural network layer and composed of elements that are each an identity matrix to generate a first tensor, and performing operations to cause each element of the first tensor to be divided by a number of elements in the window of the average pooling neural network layer to generate an initial output tensor.

Abstract: Methods, systems, and computer storage media for implementing neural networks in fixed point arithmetic computing systems. In one aspect, a method includes the actions of receiving a request to process a neural network using a processing system that performs neural network computations using fixed point arithmetic; for each node of each layer of the neural network, determining a respective scaling value for the node from the respective set of floating point weight values for the node; and converting each floating point weight value of the node into a corresponding fixed point weight value using the respective scaling value for the node to generate a set of fixed point weight values for the node; and providing the sets of fixed point floating point weight values for the nodes to the processing system for use in processing inputs using the neural network.

Abstract: Methods and systems for receiving a request to implement a neural network comprising an average pooling layer on a hardware circuit, and in response, generating instructions that when executed by the hardware circuit, cause the hardware circuit to, during processing of a network input by the neural network, generate a layer output tensor that is equivalent to an output of the average pooling neural network layer by performing a convolution of an input tensor to the average pooling neural network layer and a kernel with a size equal to a window of the average pooling neural network layer and composed of elements that are each an identity matrix to generate a first tensor, and performing operations to cause each element of the first tensor to be divided by a number of elements in the window of the average pooling neural network layer to generate an initial output tensor.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for depth concatenation using a matrix computation unit. One of the methods includes: receiving a request to process network inputs to a neural network using an integrated circuit, the neural network comprising a depth concatenation neural network layer; and generating instructions that, when executed by the integrated circuit, cause the integrated circuit to performing operations comprising: for each spatial location in a first input tensor to the depth concatenation layer and a second input tensor to the depth concatenation layer: multiplying, using the matrix computation unit, a second depth vector for the spatial location by a shift weight matrix for the depth concatenation layer to generate a shifted second depth vector; and adding the shifted second depth vector and a first input depth vector for the spatial location to generate a concatenated depth vector.

Abstract: Methods and systems for receiving a request to implement a neural network comprising an average pooling layer on a hardware circuit, and in response, generating instructions that when executed by the hardware circuit, cause the hardware circuit to, during processing of a network input by the neural network, generate a layer output tensor that is equivalent to an output of the average pooling neural network layer by performing a convolution of an input tensor to the average pooling neural network layer and a kernel with a size equal to a window of the average pooling neural network layer and composed of elements that are each an identity matrix to generate a first tensor, and performing operations to cause each element of the first tensor to be divided by a number of elements in the window of the average pooling neural network layer to generate an initial output tensor.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for depth concatenation using a matrix computation unit. One of the methods includes: receiving a request to process network inputs to a neural network using an integrated circuit, the neural network comprising a depth concatenation neural network layer; and generating instructions that, when executed by the integrated circuit, cause the integrated circuit to performing operations comprising: for each spatial location in a first input tensor to the depth concatenation layer and a second input tensor to the depth concatenation layer: multiplying, using the matrix computation unit, a second depth vector for the spatial location by a shift weight matrix for the depth concatenation layer to generate a shifted second depth vector; and adding the shifted second depth vector and a first input depth vector for the spatial location to generate a concatenated depth vector.

Abstract: Methods, systems, and computer storage media for implementing neural networks in fixed point arithmetic computing systems. In one aspect, a method includes the actions of receiving a request to process a neural network using a processing system that performs neural network computations using fixed point arithmetic; for each node of each layer of the neural network, determining a respective scaling value for the node from the respective set of floating point weight values for the node; and converting each floating point weight value of the node into a corresponding fixed point weight value using the respective scaling value for the node to generate a set of fixed point weight values for the node; and providing the sets of fixed point floating point weight values for the nodes to the processing system for use in processing inputs using the neural network.

Abstract: Methods and systems for receiving a request to implement a neural network comprising an average pooling layer on a hardware circuit, and in response, generating instructions that when executed by the hardware circuit, cause the hardware circuit to, during processing of a network input by the neural network, generate a layer output tensor that is equivalent to an output of the average pooling neural network layer by performing a convolution of an input tensor to the average pooling neural network layer and a kernel with a size equal to a window of the average pooling neural network layer and composed of elements that are each an identity matrix to generate a first tensor, and performing operations to cause each element of the first tensor to be divided by a number of elements in the window of the average pooling neural network layer to generate an initial output tensor.

Abstract: Methods and systems for receiving a request to implement a neural network comprising an average pooling layer on a hardware circuit, and in response, generating instructions that when executed by the hardware circuit, cause the hardware circuit to, during processing of a network input by the neural network, generate a layer output tensor that is equivalent to an output of the average pooling neural network layer by performing a convolution of an input tensor to the average pooling neural network layer and a kernel with a size equal to a window of the average pooling neural network layer and composed of elements that are each an identity matrix to generate a first tensor, and performing operations to cause each element of the first tensor to be divided by a number of elements in the window of the average pooling neural network layer to generate an initial output tensor.

Abstract: Methods and systems for receiving a request to implement a neural network comprising an average pooling layer on a hardware circuit, and in response, generating instructions that when executed by the hardware circuit, cause the hardware circuit to, during processing of a network input by the neural network, generate a layer output tensor that is equivalent to an output of the average pooling neural network layer by performing a convolution of an input tensor to the average pooling neural network layer and a kernel with a size equal to a window of the average pooling neural network layer and composed of elements that are each an identity matrix to generate a first tensor, and performing operations to cause each element of the first tensor to be divided by a number of elements in the window of the average pooling neural network layer to generate an initial output tensor.

Abstract: Methods and systems for receiving a request to implement a neural network comprising an average pooling layer on a hardware circuit, and in response, generating instructions that when executed by the hardware circuit, cause the hardware circuit to, during processing of a network input by the neural network, generate a layer output tensor that is equivalent to an output of the average pooling neural network layer by performing a convolution of an input tensor to the average pooling neural network layer and a kernel with a size equal to a window of the average pooling neural network layer and composed of elements that are each an identity matrix to generate a first tensor, and performing operations to cause each element of the first tensor to be divided by a number of elements in the window of the average pooling neural network layer to generate an initial output tensor.