Abstract: A computation unit for performing a computation of a neural network layer is disclosed. A number of processing element (PE) units are arranged in an array. First input values are provided in parallel in an input dimension of the array during a first processing period, and a second input values are provided in parallel in the input dimension during a second processing period. Computations are performed by the PE units based on stored weight values. An adder coupled to the first set of PE units generates a first sum of results of the computations by the first set of PE units during the first processing cycle, and generates a second sum of results of the computations during the second processing cycle. A first accumulator coupled to the first adder stores the first sum, and further shifts the first sum to a second accumulator prior to storing the second sum.

Abstract: A system and method for a software update management system which provides control over the distribution of software updates and increases payload distribution efficiency by providing a dynamically controlled next check-in frequency to each endpoint of the system. The system may use information provided from other services stored in a service database to determine the next check-in frequency. The update management system may further incorporate machine learning to optimize the next check-in frequency for each endpoint.

Abstract: A method for generating an automated test configured to test a system under test. The system under test has a plurality of operational states, at least one operational state having one or more executable actions associated therewith operable to execute predetermined operations and/or transition the system under test between operational states. The method includes the steps of: defining a model of the system under test comprising a plurality of model states; defining one or more selectable model actions; associating one or more test description sections with one or more model actions; selecting a sequence of model actions; and utilising the test description sections associated with the selected sequence of model actions to define a sequence of operation commands for execution on the system under test as an automated test.

Abstract: There is provided a method for defining an automated test configured, when executed, to test a system under test comprising one or more computer programs being executed on one or more computer devices. The system under test comprises a plurality of operational states, at least one operational state having one or more executable actions associated therewith operable to execute predetermined operations and/or transition the system under test between operational states.

Abstract: Embodiments of the present subject matter relate to implementation of a decentralized platform for deploying at least one AI (Artificial Intelligence) model. A SaaS interface is configured in a network for publishing the at least one AI model and a genesis block of a blockchain is created for the at least one AI model. Upon achieving a consensus amongst a plurality of validators in the network, the blockchain is updated and broadcasted to the network participants. Further, the at least one AI model is used as a Non Fungible Token (NFT) on the blockchain. A smart contract is generated based on a corresponding differential creativity score for the AI model The smart contract is subsequently used for creating a new block on the public blockchain.

Abstract: An example apparatus to perform a convolution on an input tensor includes a parameters generator to: generate a horizontal hardware execution parameter for a horizontal dimension of the input tensor based on a kernel parameter and a layer parameter; and generate a vertical hardware execution parameter for a vertical dimension of the input tensor based on the kernel parameter and the layer parameter; an accelerator interface to configure a hardware accelerator circuitry based on the horizontal and vertical hardware execution parameters; a horizontal Iterator controller to determine when the hardware accelerator circuitry completes the first horizontal iteration of the convolution; and a vertical Iterator controller to determine when the hardware accelerator circuitry completes the first vertical iteration of the convolution.

Abstract: This technical solution generally refers to computing systems based on biological models, and specifically to ways of training deep neural networks based on distributions of the pairwise similarity measures.

Abstract: Methods and systems for providing distributed user agent information updating. One system includes a data processing server configured to receive an update event associated with an update to an active version of user agent information. The active version of the user agent information is actively accessible for request enrichment. The data processing server is also configured to generate an updated version of the user agent information according to the update and replace the active version with the updated version in storage. In response to the storage of the updated version, the updated version is actively accessible for request enrichment.

Abstract: A method of generation of a unified view of a software development process that spans a plurality of distinct software development applications having distinct software models is provided. The method includes enhancing a software development application selected as primary to provide the unified view of the software development process, or alternatively using a data warehouse to provide the unified view of the software development process via reports and dashboards created based on a unified software model. The method provides end to end visibility, auditability, and support for calculation of one or more flow metrics, not only for the software development process, but also for a set of stages, individual stages and sub-stages of the software development process. The method also provides an ability to relate the one or more flow metrics to business outcomes.

Abstract: Application generation and integration application generation methods are disclosed. An application generation function includes automatically providing a set of software artifacts to an application developer that may be used as an aid to providing an application with augmented capabilities for an end-user. The augmented capabilities configured to adhere to certain rules regarding certification criteria and requirements of a multi-tenant cloud infrastructure. Generated applications and plug-in capabilities may further be required to meet certification criteria applicable to an application store hosted within a cloud infrastructure.

Abstract: A vehicular software update apparatus is used to a vehicle to update software stored in an electronic control unit mounted on the vehicle. The vehicular software update apparatus includes a low power communication device configured to perform a wide area wireless communication with a low power consumption. The low power communication device is operated in an update confirmation state including a state where neither power generation in the vehicle nor power supplying to the vehicle is performed, and is caused to download software update information which is information necessary to update the software, when the software update information, is provided by the server.

Abstract: The disclosed embodiments provide a system for processing data. During operation, the system obtains a function call for a function that calculates an attribute associated with a machine learning model. For each argument of the function call, the system identifies a parameter type of the argument, wherein the parameter type represents a type of data used with the machine learning model. The system also obtains a value accessor for retrieving features specific to the parameter type and obtains a value represented by the argument using the value accessor. The system then calculates the attribute by applying the function to the value and uses the attribute to execute the machine learning model.

Abstract: A method for displaying differences between a first executable dataflow graph and a second executable dataflow graph includes comparing a specification of the first executable dataflow graph and a specification of the second executable dataflow graph, including at least one of identifying a particular node or link of the first dataflow graph that does not correspond to any node or link of the second dataflow graph; and identifying a first node or link of the first dataflow graph that corresponds to a second node or link of the second dataflow graph, and identifying a difference between the first node or link and the second node or link. The method includes formulating and displaying a graphical representation of at least some of the nodes or links of the first dataflow graph or the second dataflow graph, the graphical representation including a graphical indicator of at least one of the identified particular node or link the identified difference between the first node or link and the second node or link.

Abstract: Techniques for adaptive thresholding are provided. First and second data points are received. A plurality of data points are identified, where the plurality of data points corresponds to timestamps associated with the first and second data points. At least one cluster is generated for the plurality of data points based on a predefined cluster radius. Upon determining that the first data point is outside of the cluster, the first data point is labeled as anomalous. A predicted value is generated for the second data point, based on processing data points in the cluster using a machine learning model, and a deviation between the predicted value and an actual value for the second data point is computed. Upon determining that the deviation exceeds a threshold, the second data point is labeled as anomalous. Finally, computing resources are reallocated, based on at least one of the anomalous data points.

Abstract: Systems and methods for executing compiled code having parallel code fragments is provided. One method includes storing executable code having a plurality of parallel code fragments, each of the plurality of parallel code fragments representing alternative executable paths through a code stream. The method further includes determining a code level supported by a processor executable at a computing system, the processor executable supporting a hosted computing environment. The method also includes translating the executable code into machine-readable code executable by a processor of the computing system. Translating the executable code includes selecting a code fragment from among the plurality of parallel code fragments for execution based on the code level supported by the processor executable. The method includes executing the machine-readable code within the hosted computing environment.

Abstract: A method and system for constructing a lightweight container-based user environment (CUE), and a medium, the method including: preparing, by a main process, for communication, cloning a child process, and then becoming a parent process; elevating, by the child process, permission, executing namespace isolation, and cloning a grandchild process, and setting, by the parent process, cgroups for the grandchild process; and setting, by the grandchild process, permission of the grandchild process to execute a command and a file, preparing an overlay file system, setting a hostname, restricting permission, and executing an initialization script to start the container. Multiple users are allowed to customize their own environments, enabling the users to customize their environments more flexibly, achieving privacy isolation, and making it easier and more secure to update a system. Therefore, it is particularly applicable to a high-performance computing cluster.

Abstract: Embodiments provide systems and methods for implementing a customizable enterprise automation test framework. A workflow definition, page structure definition, and function definition for an automated test of an enterprise website can be received. A hybrid script parser can parse the workflow definition, page structure definition, and function definition to generate a hybrid script for the automated test. An automation tool parser can parse the hybrid script to generate an output for an automation tool. Based on the output from the automation tool parser, a runtime script can be generated that is executed by the automation tool to generate results for the automated test, where the automation tool implements the steps of the one or more workflows on the plurality of web pages of the enterprise web site to generate the results for the automated test.

Abstract: Features are disclosed for dynamically provisioning an application stack using a set of infrastructure resources. A computing device can receive an infrastructure template from an administrative device. Based on the infrastructure template, the computing device can determine a schema and an infrastructure as code. Using the schema, the computing device can receive a specification file from a developer device or an administrative device. The specification file can be a service specification file or an environment specification file. The computing device can inject the values of the specification file into the infrastructure as code. The computing device can use the injected infrastructure as code to generate and deploy the application stack to the developer device.

Abstract: There is provided a system and method of resource verification for an application, the method comprising: obtaining a checklist of resources required for running the application, wherein the checklist is generated by: obtaining source code and an executable file of the application; running the executable file, and monitoring resources accessed by the application during execution thereof and corresponding access type of each resource, giving rise to a first list of resources; performing static code analysis of the source code including searching the source code for one or more specific code patterns indicative of respective resources to be accessed and corresponding access types thereof, and generating a second list of resources; and mapping the first list and second list of resources to generate the checklist of resources. The method further comprises automatically verifying the checklist of resources, giving rise to a verification result.

Abstract: Disclosed is an application cleaning method and device, a storage medium and an electronic device. In the embodiments of the present disclosure, the method involves: collecting multi-dimensional features of an application as samples, and constructing a sample set of the application; according to information gain ratios of the features with regard to sample classification, carrying out sample classification on the sample set so as to construct a decision tree model of the application; according to a prediction time, collecting corresponding multi-dimensional features of the application as prediction samples; and according to the prediction samples and the decision tree model, predicting whether the application can be cleaned.