Abstract: The high memory bandwidth demand of sparse embedding layers continues to be a critical challenge in scaling the performance of recommendation models. A lightweight and scalable graph-based algorithm-system co-design framework is proposed to significantly improve the embedding layer performance of recommendation models. This framework includes a novel item co-occurrence graph that scalably records item co-occurrences. Additionally, a new system-aware graph clustering algorithm is presented to find frequently accessed item combinations of arbitrary lengths to compute and memorize their partial sums. High-frequency partial sums are stored in a software-managed cache space to reduce memory traffic and improve the throughput of computing sparse features.

Abstract: A method of designing a radar, comprises receiving data pertaining to a set of reflected signals received from a distribution of objects by a respective set of receiving antennas at a respective set of locations, and feeding the data and the locations as training data to a machine learning procedure. The machine learning procedure calculates, simultaneously, a set of learned antenna locations and a set of learned parameters associating the signals with the objects, thereby providing a trained machine learning procedure parametrized by the set of learned parameters.

Abstract: There is provided an ultrasound device comprising: ultrasound transducer(s), memory storing code, hardware processor(s) coupled to the ultrasound transducer(s) and the memory for executing stored code comprising: activating the transducer(s) for simultaneously transmitting at least one line with a predefined focus width, and receiving an indication of a plurality of narrow-focused received lines, inputting the indication of the narrow-focused received lines into a convolutional neural network (CNN) trained on pairs of ultrasound imaging data capturing rapid motion of the certain tissue of sample patient(s), including a single-line ultrasound imaging data captured based on a single line in response to a transmitted single narrow-focused pulse, and a multiple-line ultrasound imaging data captured based on receiving narrow-focused lines in response to simultaneously transmitted at least one line, outputting adjusted narrow-focused received lines, and computing an adjusted ultrasound image according to the adjust

Abstract: There is provided an ultrasound device comprising: ultrasound transducer(s), memory storing code, hardware processor(s) coupled to the ultrasound transducer(s) and the memory for executing stored code comprising: activating the transducer(s) for simultaneously transmitting at least one line with a predefined focus width, and receiving an indication of a plurality of narrow-focused received lines, inputting the indication of the narrow-focused received lines into a convolutional neural network (CNN) trained on pairs of ultrasound imaging data capturing rapid motion of the certain tissue of sample patient(s), including a single-line ultrasound imaging data captured based on a single line in response to a transmitted single narrow-focused pulse, and a multiple-line ultrasound imaging data captured based on receiving narrow-focused lines in response to simultaneously transmitted at least one line, outputting adjusted narrow-focused received lines, and computing an adjusted ultrasound image according to the adjust