Nvidia Corporation

Greg Heinrich

Gary D Hicok

Michael Brian Cox

John R Zedlewski

Rev Lebaredian

Tony Tamasi

Clement Farabet

Claire Delaunay

David Auld

Matthew Campbell

Zachary Taylor

Mark Daly

Parthasarathy Sriram

Anil Ubale

Tayfun Gurel

Curtis A Beeson

Robert Riglar

Frederic Bossy

Curtis Beeson

Varun Praveen

Philippe Guasch

Input layers of an element-wise operation in a neural network can be pruned such that the shape (e.g., the height, the width, and the depth) of the pruned layers matches. A pruning engine identifies all of the input layers into the element-wise operation. For each set of corresponding neurons in the input layers, the pruning engine equalizes the metrics associated with the neurons to generate an equalized metric associated with the set. The pruning engine prunes the input layers based on the equalized metrics generated for each unique set of corresponding neurons.

Pruning neural networks that include element-wise operations

In various examples, physical sensor data may be generated by a vehicle in a real-world environment. The physical sensor data may be used to train deep neural networks (DNNs). The DNNs may then be tested in a simulated environment—in some examples using hardware configured for installation in a vehicle to execute an autonomous driving software stack—to control a virtual vehicle in the simulated environment or to otherwise test, verify, or validate the outputs of the DNNs. Prior to use by the DNNs, virtual sensor data generated by virtual sensors within the simulated environment may be encoded to a format consistent with the format of the physical sensor data generated by the vehicle.

Training, testing, and verifying autonomous machines using simulated environments

In various examples, physical sensor data may be generated by a vehicle in a real-world environment. The physical sensor data may be used to train deep neural networks (DNNs). The DNNs may then be tested in a simulated environment&#x2014;in some examples using hardware configured for installation in a vehicle to execute an autonomous driving software stack&#x2014;to control a virtual vehicle in the simulated environment or to otherwise test, verify, or validate the outputs of the DNNs. Prior to use by the DNNs, virtual sensor data generated by virtual sensors within the simulated environment may be encoded to a format consistent with the format of the physical sensor data generated by the vehicle.

Computer system, method and computer program product for scheduling IPC activities are disclosed. In one embodiment, the computer system includes first processor and second processors that communicate with each other via IPC activities. The second processor may operate in a first mode in which the second processor is able to process IPC activities, or a second mode in which the second processor does not process IPC activities. Processing apparatus associated with the first processor identifies which of the pending IPC activities for communicating from the first processor to the second processor are not real-time sensitive, and schedules the identified IPC activities for communicating from the first processor to the second processor by delaying some of the identified IPC activities to thereby group them together. The grouped IPC activities are scheduled for communicating to the second processor during a period in which the second processor is continuously in the first mode.

Power-efficient inter processor communication scheduling

In an aspect there is provided a method of moving a processor of a mobile device from a low-power state for conserving power to an active mode for processing signals. The mobile device is configured to receive regularly scheduled signals. The method comprises, for each of multiple operating states of the mobile device determining a restore time associated with the operating state of the mobile device and storing each determined restore time in association with its operating state. The method further comprises detecting a current operating state of the mobile device and using the determined restore time for that state to set a trigger time to control movement of the processor of the mobile device to enter the active mode from the low-power mode in time to process the scheduled signals.

Low power mode exit latency predictor for real-time asymmetric multiprocessor systems

An apparatus comprising: a first transceiver arranged to communicate over a wireless network, the first transceiver comprising a first clock; and a second transceiver arranged to communicate other than by said wireless network, the second transceiver comprising a second clock. The second sends a request signal to the first transceiver. In response, the first transceiver transitions from a first mode to a second mode and provides to the second transceiver a response signal for calibrating the second clock relative to the first clock. In the first mode the first transceiver performs zero or more calibrations of the first clock relative to the wireless network, and in the second mode the first transceiver performs at least one additional calibration of the first clock relative to the wireless network, the response signal being based on the at least one additional calibration.

Reference clock calibration

Growing community of inventors

Sophia Antipolis, France

Greg Heinrich