Abstract: Approaches in accordance with various embodiments allow for zero-touch enrollment of devices with respective manager systems. In at least one embodiment, a device at startup can contact a central directory service (CDS) for information about an associated manager. The CDS can authenticate the device using device information included in the request, and can send a challenge token to the device in response. The challenge token can include information for the manager, protected with multiple layers of security that should only be able to be decrypted by the authenticated device. The device can decrypt this challenge token to determine the manager information, and can convert this challenge token to a bearer token. The device can then send a request to the determined manager that includes the bearer token, which the manager can use to authenticate the device. The manager can then send the device appropriate configuration information.

Abstract: A neural network-based rendering technique increases temporal stability and image fidelity of low sample count path tracing by optimizing a distribution of samples for rendering each image in a sequence. A sample predictor neural network learns spatio-temporal sampling strategies such as placing more samples in dis-occluded regions and tracking specular highlights. Temporal feedback enables a denoiser neural network to boost the effective input sample count and increases temporal stability. The initial uniform sampling step typically present in adaptive sampling algorithms is not needed. The sample predictor and denoiser operate at interactive rates to achieve significantly improved image quality and temporal stability compared with conventional adaptive sampling techniques.

Abstract: Various techniques for accelerating Smith-Waterman sequence alignments are provided. For example, threads in a group of threads are employed to use an interleaved cell layout to store relevant data in registers while computing sub-alignment data for one or more local alignment problems. In another example, specialized instructions that reduce the number of cycles required to compute each sub-alignment score are utilized. In another example, threads are employed to compute sub-alignment data for a subset of columns of one or more local alignment problems while other threads begin computing sub-alignment data based on partial result data received from the preceding threads. After computing a maximum sub-alignment score, a thread stores the maximum sub-alignment score and the corresponding position in global memory.

Abstract: Techniques to generate driving scenarios for autonomous vehicles characterize a path in a driving scenario according to metrics such as narrowness and effort. Nodes of the path are assigned a time for action to avoid collision from the node. The generated scenarios may be simulated in a computer.

Abstract: Many times, users play games to certain places in the game and decide to save the game and resume it later. When users resume the game, they have to wait for it to load, wade through menus to select the saved game, and wait again for the game to resume. The introduced technique directly suspends and resume games that are rendered on a cloud server based on user initiated suspend and resume control commands. As the games are directly suspended and resumed, the users can skip past menu and screen loading. The introduced technique can track game information of multiples users in cloud tokens, and can move these cloud tokens from one memory location to another based on the user initiated suspend and resume control commands.

Abstract: Apparatuses, systems, and techniques are presented to perform monocular view synthesis of a dynamic scene. Single and multi-view depth information can be determined for a collection of images of a dynamic scene, and a blender network can be used to combine image features for foreground, background, and missing image regions using fused depth maps inferred form the single and multi-view depth information.

Abstract: Users often desire to capture certain images from an application. Existing methods of capturing images can result in low-resolution images due to limitations of the display device providing the images. This disclosure provides a method of capturing higher resolution images from source images. Techniques are also disclosed to reduce the storage size associated with the higher resolution images. Through capturing low-resolution versions of the same source images, image effects can be captured and applied to the higher resolution images where those image effects may be altered or missing. Frequency spectrum combination can be used to combine the low-resolution image data and the higher resolution image data. The higher resolution images can be processed using a segmentation scheme, such as tiling, without reducing or limiting the image effects.

Abstract: This disclosure provides an integrated circuit device that includes a RDL that is interlocked with a bump (or “pillar”). The interlocked interface provides the contact RDL-bump interface with increased structural stability that can better withstand the thermal stresses associated with high performance devices IC devices. The interlock structure mitigates crack/delamination that occurs at the RDL-bump interface in large IC chips that are generally subjected to higher stresses during operation.

Abstract: In various examples, sensor data may be received that represents a field of view of a sensor of a vehicle located in a physical environment. The sensor data may be applied to a machine learning model that computes both a set of boundary points that correspond to a boundary dividing drivable free-space from non-drivable space in the physical environment and class labels for boundary points of the set of boundary points that correspond to the boundary. Locations within the physical environment may be determined from the set of boundary points represented by the sensor data, and the vehicle may be controlled through the physical environment within the drivable free-space using the locations and the class labels.

Abstract: In various examples, sensor data may be adjusted to represent a virtual field of view different from an actual field of view of the sensor, and the sensor data—with or without virtual adjustment—may be applied to a stereographic projection algorithm to generate a projected image. The projected image may then be applied to a machine learning model—such as a deep neural network (DNN)—to detect and/or classify features or objects represented therein.

Abstract: A machine memory includes multiple memory cells. Word lines, each with at least one word line driver, are coupled to the memory cells along rows. The word line drivers of at least some adjacent pairs of the word lines are coupled together by a pull-down transistor, in a manner that reduces read disturb of the memory cells.

Abstract: In various examples, a deep neural network(s) (e.g., a convolutional neural network) may be trained to detect moving and stationary obstacles from RADAR data of a three dimensional (3D) space. In some embodiments, ground truth training data for the neural network(s) may be generated from LIDAR data. More specifically, a scene may be observed with RADAR and LIDAR sensors to collect RADAR data and LIDAR data for a particular time slice. The RADAR data may be used for input training data, and the LIDAR data associated with the same or closest time slice as the RADAR data may be annotated with ground truth labels identifying objects to be detected. The LIDAR labels may be propagated to the RADAR data, and LIDAR labels containing less than some threshold number of RADAR detections may be omitted. The (remaining) LIDAR labels may be used to generate ground truth data.

Abstract: A deep neural network(s) (DNN) may be used to detect objects from sensor data of a three dimensional (3D) environment. For example, a multi-view perception DNN may include multiple constituent DNNs or stages chained together that sequentially process different views of the 3D environment. An example DNN may include a first stage that performs class segmentation in a first view (e.g., perspective view) and a second stage that performs class segmentation and/or regresses instance geometry in a second view (e.g., top-down). The DNN outputs may be processed to generate 2D and/or 3D bounding boxes and class labels for detected objects in the 3D environment. As such, the techniques described herein may be used to detect and classify animate objects and/or parts of an environment, and these detections and classifications may be provided to an autonomous vehicle drive stack to enable safe planning and control of the autonomous vehicle.

Abstract: The disclosure provides using test processors to provide a more flexible solution compared to the existing DFX blocks that are used for controlling test networks in chips. The test processors provide a highly flexible solution since programming of the test processors can be changed at any time; even after manufacturing, and can support practically an unlimited number of core chips in any configuration. The high flexibility provided via the test processors can reduce engineering effort needed in design and verification, accelerate schedules, and may prevent additional tapeouts in case of DFX design bugs. By making debug and diagnosis easier by providing an opportunity to change debug behavior as needed, the time-to-market timeline can be accelerated. Accordingly, the disclosure provides a chip with a test processor, a multi-chip processing system with a test processor, and a method of designing a chip having a test processor.

Abstract: A protective shroud includes a top plate, a first side plate that is adapted to be disposed proximate a first edge region of a plurality of cooling fins of a heat exchanger for an integrated circuit, and a second side plate that is adapted to be disposed proximate a second edge region of the plurality of cooling fins.

Abstract: A packed error correction code (ECC) technique opportunistically embeds ECC check-bits with compressed data. When compressed, the data is encoded in fewer bits and is therefore fragmented when stored or transmitted compared with the uncompressed data. The ECC check-bits may be packed with compressed data at “source” points. The check-bits are transmitted along with the compressed data and, at any “intermediate” point between the source and a “destination” the check-bits may be used to detect and correct errors in the compressed data. In contrast with conventional systems, packed ECC enables end-to-end coverage for sufficiently-compressed data within the processor and also externally. While storage circuitry typically is protected by structure-specific ECC, protection is also beneficial for data as it is transmitted between processing and/or storage units.

Abstract: In various examples, a path perception ensemble is used to produce a more accurate and reliable understanding of a driving surface and/or a path there through. For example, an analysis of a plurality of path perception inputs provides testability and reliability for accurate and redundant lane mapping and/or path planning in real-time or near real-time. By incorporating a plurality of separate path perception computations, a means of metricizing path perception correctness, quality, and reliability is provided by analyzing whether and how much the individual path perception signals agree or disagree. By implementing this approach—where individual path perception inputs fail in almost independent ways—a system failure is less statistically likely. In addition, with diversity and redundancy in path perception, comfortable lane keeping on high curvature roads, under severe road conditions, and/or at complex intersections, as well as autonomous negotiation of turns at intersections, may be enabled.

Abstract: According to one or more embodiments, operations may comprise obtaining a first point cloud. The operations also comprise performing segmentation of the first point cloud, the segmentation generating one or more clusters of points of the point cloud. The operations also comprise determining, for each respective cluster of the plurality of clusters, a respective geometric feature of a corresponding object that corresponds to the respective cluster. The operations also comprise obtaining a second point cloud. The operations also comprise assigning a plurality of weights that comprises assigning a respective weight to each respective cluster based on the respective geometric feature that corresponds to the respective cluster. The operations also comprise obtaining a second point cloud and aligning the first point cloud with the second point cloud based on the plurality of weights.

Abstract: In various examples, historical trajectory information of objects in an environment may be tracked by an ego-vehicle and encoded into a state feature. The encoded state features for each of the objects observed by the ego-vehicle may be used—e.g., by a bi-directional long short-term memory (LSTM) network—to encode a spatial feature. The encoded spatial feature and the encoded state feature for an object may be used to predict lateral and/or longitudinal maneuvers for the object, and the combination of this information may be used to determine future locations of the object. The future locations may be used by the ego-vehicle to determine a path through the environment, or may be used by a simulation system to control virtual objects—according to trajectories determined from the future locations—through a simulation environment.

Abstract: Diagnostics and boot up for AV hardware and software of a computer system of an autonomous vehicle may be performed based at least on receiving a shutdown or power off indication, then a computing state of the computer system may be suspended with the computer system entering a low-power mode. The suspended computing state can be rapidly restored without requiring a reboot and diagnostics for key-on. To ensure the integrity of the saved computing state, the computer system may exit the low-power mode, rerun the diagnostics, reload the programs, and then reenter the low-power mode. Restoring the suspended computing state may be triggered by a user inserting an ignition key, pressing a button to turn on the vehicle, opening a door to the vehicle, remotely unlocking the vehicle, remotely starting the vehicle, etc.