Abstract: A circuit may enable communication between a primary device and one or more secondary devices. The communication may utilize a Universal Asynchronous Receiver Transmitter (UART) protocol. In operation, the primary device may require information on the baud rate of the secondary device. The UART may operate in an inverted polarity mode, and this inverted polarity may be interpreted by the secondary device as a request to enter an auto-baud detection mode. Using the inverted polarity mode to enter the auto-baud detection mode may prevent excessive delays in the UART communication and may prevent the need for additional pins to implement the auto-baud detection mode.

Abstract: In various embodiments, a system can: access a failure image on which a first model has inaccurately performed an inferencing task; train, on a set of dummy images, a second model to learn a visual variety of the failure image, based on a loss function having a first term and a second term, the first term quantifying visual content dissimilarities between the set of dummy images and outputs predicted during training by the second model, and the second term quantifying, at a plurality of different image scales, visual variety dissimilarities between the failure image and the outputs predicted during training by the second model; and execute the second model on each of a set of training images on which the first model was trained, thereby yielding a set of first converted training images that exhibit the visual variety of the failure image.

Abstract: Systems and methods are provided for optimizing a driver's trajectory around a race track. The system can receive driver data representing driver inputs to a vehicle. A vehicle state can be determined for at least one location on the race track based on the received driver data. Expert driver data can be determined representing inputs of an expert driver at the determined vehicle state for the at least one location. The system can predict how the expert driver would proceed from each current vehicle state based on the expert driver data and map the current vehicle state at the at least one location to a recommended trajectory from each current vehicle state based on predicting how the expert driver would proceed.

Abstract: According to one embodiment, a method, computer system, and computer program product for mitigating a camera failure during volumetric capture is provided. The embodiment may include monitoring respective status of functionality (SoF) array data structures of a set of cameras tasked with capturing volumetric video of a scene. The set of cameras comprises cameras designated as backup cameras and cameras designated for active volumetric capture. The embodiment may include detecting a failed camera of the cameras designated for active volumetric capture based on evaluation of an SoF array data structure of the failed camera. In response to detecting the failed camera, the embodiment may include determining whether a backup camera of the cameras designated as backup cameras is available to replace the failed camera. In response to determining the backup camera is available, the embodiment may include replacing the failed camera with the backup camera.

Abstract: Apparatuses, systems, and techniques to perform computational operations in response to one or more compute uniform device architecture (CUDA) programs. In at least one embodiment, one or more computational operations are to cause one or more other computational operations to wait until a portion of matrix multiply-accumulate (MMA) operations have been performed.

Abstract: A data protection system uses machine learning to detect the cyber-attacks on a data protection system in advance to notify the user of possible attacks and also instigate any counter attacks to the best possible extent. The system trains a support vector machine model (SVM) to recognize a malware, or other type of attack before it comes into action against the system. This model learns the parameters of hazardous files or code to prepare the best model of attributes of such files to help block the malware proactively. It uses several independent variables as features to gain more accuracy to the threat detection. Some of the parameters include: rate of data change (Drastic/High/Low), attack vulnerability history, resource usage history, performance metrics, application hit ratio, and the like.

Abstract: A method includes receiving a learning requirement for a user. A preliminary learning recommendation is generated based on the learning requirement of the user and learning histories of a plurality of users. The preliminary learning recommendation is filtered to generate a filtered learning recommendation. A problem of finding an improved learning recommendation from the filtered learning recommendation is mapped to a quantum Ising model. A quantum annealing process is performed one or more times to determine one or more final quantum states from an initial quantum state. One or more energies of the quantum Ising model are determined for the one or more final quantum states. A minimum energy is determined among the determined energies. A target quantum state that corresponds to the minimum energy is determined among the determined final quantum states. The improved learning recommendation is determined from the target quantum state and is sent to the user.

Abstract: Prioritization of VT scans can be performed using particular select gates of a memory device or memory sub-system in the absence of performing such select gate scan operations on all of the select gates of an entire memory die or of all the memory dice of a memory device or memory sub-system. A method for such prioritization of VT scans includes determining quality characteristics of a memory die and altering a threshold voltage applied to the memory die in performance of a select gate scan operation based, at least in part, on the determined quality characteristics of the memory die. Such methods can further include performing the select gate scan operation by applying signaling having the altered threshold voltage to a select gate of the memory die.

Abstract: A dynamic data policy creation process utilizes certain supervised learning processes to classify data criticality for tagged clients to provide dynamic policy definitions and process adhoc (special) backup requests. Such embodiments prevent the need to manually determine criticality and create protection policies whenever a new data object into the system. A data tagger and KNN-based classifier provide an intelligent solution to data protection ecosystems to meet the dynamic request of data objects through a dynamic backup policy creation system and method that uses certain artificial intelligence (AI) and machine learning (ML) based solutions.

Abstract: A container prioritization process helps schedule backups of containerized data based on defined attributes and historical data. Containers are classified using a KNN-based classifier based on attributes. A tagger component assigns a priority tag to each container. Containers are backed up by a backup server through a schedule based on the priority tags of the containers. New containers are automatically classified and tagged within the prioritization schedule using the KNN-based classifier.

Abstract: This disclosure provides systems, devices, apparatus, and methods, including computer programs encoded on storage media, for adaptive BVH rebuilds with biased cost functions for dynamic geometry. A graphics processor may obtain an indication of first BVH structure(s) including first nodes, where the first BVH structure(s) are representative of first geometry data for first primitives in first frame(s), where each of the first nodes is associated with first primitive(s), may detect a number of rays that intersect each of the first BVH structure(s) from direction(s) associated with the first frame(s), may update a cost function based on the number of rays and each of the direction(s), and may configure, based on the updated cost function, second BVH structure(s) including second nodes, where the second BVH structure(s) are representative of second geometry data for second primitives in second frame(s), where each of the second nodes is associated with second primitive(s).

Abstract: Systems, apparatus, articles of manufacture, and methods are disclosed to improve fan operability control using smart materials. An engine comprising an engine surface in an airflow path, a sensor positioned on the engine surface, and a smart-material-based feature positioned on the engine surface, the smart-material-based feature triggered to modify the airflow path when the sensor outputs an indication of a detected deviation from a reference value of an operating parameter of the engine.

Abstract: A data protection system utilizes certain supervised learning processes to implement a granular level prioritized protection scheme for critical data based on decision tree processing. This process provides intelligent backup protection not only at the asset or sub-asset level, but also at the molecular level to meet the desired level of priority-based operation. Such embodiments overcome disadvantages of current systems that provide this level protection only upon management by an administrator by providing automated intelligence to provide granular protection at any appropriate data object level.

Abstract: Systems and methods for providing collision-free unique versioned identifiers using a distributed ledger are provided herein. An example system may include a unique identifier provider that is configured to provide evolving, backward compatible, unique versioned identifiers from semi-structured data such as forms or claims to various in a decentralized fashion.

Abstract: The backup schedule for a group of routine or known clients and data objects is generally defined by default backup policies and priorities, and such policies and data assets demand resources of the system. The resource allocation among data assets being processed is analyzed and reallocated (tuned) as necessary to ensure that critical data assets are provided with the necessary resources for their protection operations. Embodiments use a machine learning (ML) model that would identify the resource consumption for the all the running policies to smartly allocate the optimal number of resources based on the criticality of the data object and their policies.

Abstract: A data protection system implements a Naïve Bayes classifier-based server health resiliency process that greatly helps the amount of time needed to resolve any health-based issue in the server. The Naive Bayes is an example of a simple classifier that classifies based on probabilities of problematic or potential failure causing events. This helps empower vendor applications to intelligently identify automatically resolve these flaws without the need for vendor personnel on the customer environment. Such a process uses historical cases and trains machine learning models in such a way that troubleshooting, log analysis and recommendations will be done proactively to identify root causes of issue and identify and apply available and appropriate fixes and workarounds.

Abstract: A conflict resolution component uses data change measure to resolve conflicts when data assets with the same priority tags arrive simultaneously for protection processing. The data change measure quantifies the extent of modifications or updates made to the asset since a last backup. Assets with a higher data change measure are assigned a higher priority and processed ahead of others with the same priority tag. This reprioritization ensures that backup objects with more significant data changes are handled first. Such a system overcomes the issues associated with present methods backup queueing methods including random scheduling of data having the same priority tags or classifications.

Abstract: A container load balancer process helps schedule backups of containerized data based on defined attributes and historical data. Containers are classified using a KNN-based classifier based on attributes. A tagger component assigns a priority tag to each container. A monitor monitors an assignment of backup tasks to proxies for the backing up step, and a load balancer determines if the assignment distributes backup loads within a defined performance tolerance, and adjusts the assignment if not. A backup server then backs up the container data in an order determined by the classifying and the assignment or adjusted assignment.

Abstract: A dynamic replication service using a data change metric to select an optimum cloning method that reduces latency of data copying. A model is trained using historical data of backup operations of the saveset to establish past data change metrics for corresponding replication services processing the saveset. The best cloning method for the replication service is selected by using a calculated data change rate of the data saveset, as expressed as a number of bytes changed per unit of time, from among a plurality of different cloning methods based on the data change rate. The service executes the selected cloning method for the replication service to copy the data for storage or further processing.

Abstract: Systems and techniques are described herein for network organization. For instance, a process can include receiving, from a second sensing apparatus of the self-organizing network, an indication of a first task. The process can further include retrieving a first power profile associated with the first task, determining a current battery level of the at least one battery, predicting a future battery level of the at least one battery based on the current battery level, and the first power profile associated with the first task, and transmitting, to another sensing apparatus, the predicted future battery level.