Abstract: Techniques for joining a device of a third user to a communication between a device of a first user and a device of a second user are described herein. For instance, two or more users may utilize respective computing devices to engage in a telephone call, a video call, an instant-messaging session, or any other type of communication in which the users communicate with each other audibly and/or visually. In some instances, a first user of the two users may issue a voice command requesting to join a device of a third user to the communication. One or more computing devices may recognize this voice command and may attempt to join a device of a third user to the communication.

Abstract: Aspects of a data environment, such as the creation, provisioning, and management of data stores and instances, are managed using a separate control environment. A user can call into an externally-facing interface of the control environment, the call being analyzed to determine actions to be performed in the data environment. A monitoring component of the control plane also can periodically communicate with the data environment to determine any necessary actions to be performed, such as to recover from faults or events in the data environment. A workflow can be instantiated that includes tasks necessary to perform the action. For each task, state information can be passed to a component in the data environment operable to perform the task, until all tasks for an action are completed. Data in the data environment can be accessed directly using an externally-facing interface of the data environment, without accessing the control plane.

Abstract: An improved search system can address sparse behavioral signals. Machine learning is used to determine queries with historical data that are similar to an input query, which have limited behavioral signals. Co-acquisitions of items for query pairs are used to generate a similarity value for the pairs, which are used as labels during training of machine learning models. Multiple machine learning models are used during inference such that a subset of similar queries is determined with a less computationally expensive model and a more computationally expensive model is used to re-rank the subset of queries. Accordingly, for input queries with limited behavioral data, the search system identifies similar queries with behavioral data and their corresponding search results can be presented to a user.

Abstract: A multi-column index is generated based on an interleaving of data bits for selectivity for efficient processing of data in a relational database system. Two or more columns may be identified for inclusion in the multi-column index for a relational database table. Based, at least in part, on the interleaving of data bits for selectivity from the identified columns, a multi-column index is generated for the relational database table that provides a respective index value for each entry in the relational database table. The entries of the relational database table may then be stored according to the index values of the multi-column index.

Abstract: A system and method for authorization policy validation. A validator takes as input an authorization policy to be analyzed and a schema that specifies entity types and their attributes, types of entity parents in an entity hierarchy, and which entity types can be used with which actions. The validator checks that the policy conforms to the schema. If the check passes, then the policy is guaranteed to be free of both type errors and attribute access errors for any input that conforms to the schema.

Abstract: A rain-resistant, drainless, heated air data probe assembly may comprise a probe having a central channel, one or more peripheral channels, and a heating element. In addition, inner surfaces of the channels may include hydrophilic surfaces, and the outer surface of the probe may include hydrophobic surfaces. The hydrophobic outer surface may reduce water ingress into the channels of the probe, and the hydrophilic inner surfaces may cause dispersion of water that has entered the channels. The heating element may further cause evaporation or running off of water on the outer surface, and may cause evaporation of dispersed water within the channels, which may further eliminate the need for a drain at a terminal end of the probe assembly.

Abstract: Technologies directed to the distribution of local oscillator (LO) signals and reference clock signals to beamforming circuits of a panel with an array of antenna elements are described. A communication system includes antenna elements, beamforming circuits, driver circuitry, and diplexing circuitry with signal paths between the diplexing circuitry and the beamforming circuits. The driver circuitry outputs a first signal having a first frequency and a second signal having a second frequency lower than the first frequency. The diplexing circuitry generates a combined signal comprising the first and second signals and sends the combined signal to each beamforming circuit.

Abstract: Some embodiments provide a compiler for optimizing the implementation of a machine-trained network (e.g., a neural network) on an integrated circuit (IC). The compiler of some embodiments receives a specification of a machine-trained network including multiple layers of computation nodes and generates a graph representing options for implementing the machine-trained network in the IC. In some embodiments, the compiler also generates instructions for gating operations. Gating operations, in some embodiments, include gating at multiple levels (e.g., gating of clusters, cores, or memory units). Gating operations conserve power in some embodiments by gating signals so that they do not reach the gated element or so that they are not propagated within the gated element. In some embodiments, a clock signal is gated such that a register that transmits data on a rising (or falling) edge of a clock signal is not triggered.

Abstract: The present invention features new waveguide layouts for input, redirection (expansion), and output holograms that minimize cross talk between colors and allow all three colors to reside in a single waveguide. The use of multiple incoupling holograms that diffract different colors of light in different directions, or along different paths, through a waveguide substrate advantageously provides for a reduction of cross-talk between the colors of a holographic image. In a square-shaped design, red, green, and blue input and output holograms approximately overlay on top of each other. The green redirection hologram is laterally separated from the red and blue redirection holograms. Using this square-shape design, the light beams for the three colors are separated into two paths propagating from input to output holograms.

Abstract: Techniques for machine learning prediction using edge devices are described. In some examples a method of use includes receiving a request to load a second model onto an edge device while a first model has already been loaded on the edge device, wherein the second model and the first model share an external handle; loading at least one instance the second model into memory of the edge device; and after the second model has been loaded into memory of the edge device, directing a prediction request to the shared external handle to the second model instead of the first model.

Abstract: This disclosure describes, in part, techniques for conserving power on an electronic device. For instance, at given time intervals, the electronic device may be sending input data to a network device and receiving audio data from the network device. The electronic device may then use one or more techniques to determine when to switch from operating in a first mode, where the electronic device sends and/or receives the data, to operating in a second mode, where the electronic device ceases sending and/or receiving the data. For example, the electronic device may make the determination based on an amount of data stored in a buffer, whether the electronic device receives data, using data received from the network device, and/or the like. Based on the determination, the electronic device may switch to the second mode in order to conserve power.

Abstract: A compute channel can have multiple computational circuit blocks coupled in series to form a pipeline. The compute channel can perform a computation on an input tensor to generate an output tensor based on an instruction. When the computational does not require all of the computational circuit blocks, the throughput of the compute channel can be increased by splitting the data elements of the input tensor into multiple input data streams. The multiple input data streams are provided to respective subsets of one or more computational circuit blocks in the pipeline using bypass circuitry of the computational circuit blocks, and the computation can be performed on multiple input data streams in the respective subsets of one or more computational circuit blocks to generate multiple output data streams corresponding to the output tensor.

Abstract: Techniques for predicting an answer to a question using a machine learning model are described. In some examples, the model predicts one or more answers to the question by: predicting at least two answers to the question using a first component of the question-answer model from a set of passages, generating, using a second component of the question-answer model, at least one question for each of the predicted at least two answers, and performing roundtrip predictions until each generated question only has one answer.

Abstract: Session-specific edit recommendations may be made for editing a code file. After a code editing session is started, code file edits may be captured. A machine learning technique may be applied to learn the code edits and recommend alternative code portions for portions of the code file during the code editing session. The recommendations may be provided and accepted, or not, via an interface of a code editor application.

Abstract: A first distributed unit (DU) of a radio-based application determines that traffic associated with a particular user equipment device (UED) is to be handed over to a second DU. The first DU transmits contents of a buffer, including a data block transmitted from the UED, to the second DU using a direct network channel established between the DUs. The second DU stores the contents in a second buffer, and uses the contents to perform a retransmission operation associated with the traffic of the UED.

Abstract: To reduce the storage size of weight tensors and speed up loading of weight tensors from system memory, a compression technique can be employed to remove zero values from a weight tensor before storing the weight tensor in system memory. A sparsity threshold can be enforced to achieve a compression ratio target by forcing small weight values to zero during training. When the weight tensor is loaded from system memory, a direct memory access (DMA) engine with an in-line decompression unit can decompress the weight tensor on-the-fly. By performing the decompression in the DMA engine, expansion of the weight values back to the original weight tensor size can be carried out in parallel while other neural network computations are being performed by the processing unit.

Abstract: Systems and methods for providing vehicle software deployment plans that include one or more fallback deployment plans are disclosed. In some embodiments, a vehicle software deployment system determines a deployment plan for deploying one or more software applications one or more electronic control units (ECUs) of a vehicle. Additionally, for one or more failure scenarios, the vehicle software deployment system determines one or more respective fallback deployment plans, such as a deployment plan that assumes one of the ECUs of the vehicle has failed. A deployment plan bundle is provided to the vehicle, comprising a primary deployment plan as well as the one or more fallback deployment plans. In the event that one of the failure scenarios takes place with respect to the vehicle, a deployment agent of the vehicle automatically deploys one of the fallback deployment plans that was provided with the deployment bundle.

Abstract: Gain factors for amplifying audio data of live broadcasts from mobile devices are selected to ensure that distortions of audio data subsequently generated during the live broadcasts remain within targeted distortion ranges. When a creator initiates a live broadcast from a personal device, audio data captured by the personal device is subjected to amplification and a level of distortion of the audio data is determined. A gain factor for amplifying audio data subsequently captured during the live broadcast is calculated based on a previous gain factor, as well as distortion of the audio data resulting from hard or soft clipping, or any other basis. The gain factor may increase, decrease or remain constant, depending on the prior gain factor and the previously observed distortion.

Abstract: Techniques for altering default language, in system outputs, with language included in system inputs are described. A system may determine a word(s) in user inputs, associated with a particular user identifier, correspond to but are not identical to a word(s) in system outputs. The system may store an association between the user identifier, the word(s) in the user inputs, and the word(s) in the system outputs. Thereafter, when the system is generates a response to a user input, the system may replace the word(s), traditionally in the system outputs, with the word(s) that was present in previous user inputs. Such processing may further be tailored to a natural language intent.