Neural network processing system with multiple processors and neural network accelerator

摘要

At least one neural network accelerator 238 performs operations of the first subset of layers of the neural network on the input data set, creates an intermediate data set, and a shared memory queue 614 in shared memory 612. To save the intermediate data set. The first processor element 602 of the host computer system 402 provides input data to the neural network accelerator and signals the neural network accelerator to perform operations of the first subset of layers of the neural network on the input data set. . The second processor element 604 of the host computer system reads the intermediate data set from the shared memory queue, while the neural network accelerator performs operations of the first subset of the layers of the neural network for other input data sets, intermediate Perform operations of the second subset of the layers of the neural network on the data set, and generate an output data set.