Abstract: A machine learning accelerator (MLA) implements a machine learning network (MLN) by using data transfer instructions that coordinate concurrent data transfers between processing elements. A compiler receives a description of a machine learning network and generates the computer program that implements the MLN. The computer program contains instructions that will be run on PEs of the MLA. The PEs are connected by data transfer paths that are known to the compiler. The computations performed by the PEs may require data stored at other PEs. The compiler coordinates the data transfers to avoid conflicts and increase parallelism.

Abstract: A method or system for testing and benchmarking commits made on source code. The system extracts commits from a history of a first code history that records a sequence of commits made thereon. The extracted commits are then combined into a sequence of patches, each of which includes changes made during consecutive commits. The system also establishes a connection with a system under test (SUT) having a second code repository corresponding to a historical version of the first code repository before the extracted commits were made, and sequentially applies the sequence of the patches to the second code repository. Performance of the SUT is monitored during the application of the sequence of the patches to determine whether the SUT performs as expected after the application of each patch.

Abstract: Embodiments are disclosed for a utilization-aware approach to cluster scheduling, to address this resource fragmentation and to improve cluster utilization and job throughput. In some embodiments a resource manager at a master node considers actual usage of running tasks and schedules opportunistic work on underutilized worker nodes. The resource manager monitors resource usage on these nodes and preempts opportunistic containers in the event this over-subscription becomes untenable. In doing so, the resource manager effectively utilizes wasted resources, while minimizing adverse effects on regularly scheduled tasks.

Abstract: Techniques and solutions are described for automatically updating test code based on changes, including additions or deletions, to production code. Requests to analyze production code and a modification to production code are received. The modification is applied to the production code to provide modified production code. Test code specific for the production code is executed on the modified production code. It is determined that a test specified in the test code was failed by the modified production code. A condition responsible for test failure is identified, where the condition specifies at least one expected value or at least one expected result. A modification to update the test code is determined, where the modification includes an actual value or an actual result obtained during execution of the failed test. The modification to the test code is applied to provide modified test code.

Abstract: Systems and methods for automatic self-configuration of deployments of sets of enterprise software applications to users are disclosed. Exemplary implementations may: store information, including executable code for a set of enterprise software applications and a configuration database including deployment-specific configuration settings and corresponding setting values; monitor and quantify operations of a deployment of the set of enterprise software applications; assess whether a system-initiated modification of the deployment is likely to improve a particular monitored and quantified operation; automatically generate a system-initiated modification database based on the system-initiated modification; and modify the deployment in accordance with the system-initiated modification database.

Abstract: Techniques for processing change requests for software projects are disclosed. An example method includes generating a contributor score for a contributor of source code to a software project and saving the contributor score to a storage device. The contributor score reflects a degree to which source code submissions by the contributor are trustworthy. The method also includes receiving a change request from the contributor to change a part of the software project. The change request includes new or updated source code for the software project. The method also includes, in response to receiving the change request, retrieving the contributor score from the storage device, comparing, by a processing device, the contributor score to a threshold to identify an action to be performed in relation to the change request, and executing the action.

Abstract: A system includes a rules engine comprising a template indicative of an initial source code of a core computing system, a processor, and memory storing instructions. The instructions are configured to cause the processor to retrieve data indicative of a current source code for the core computing system, and compare the data indicative of the current source code to the template indicative of the initial source code of the core computing system. The processor may then identify one or more customizations in the current source code of the core computing system based on comparing the data to the template and quantify an amount of customization applied to the core computing system based on identifying the one or more customizations.

Abstract: Methods, apparatus, and processor-readable storage media for automated software deployment are provided herein. An example computer-implemented method includes obtaining at least one request to merge one or more code changes associated with a first branch of software code with a second branch of software code; performing one or more tests to validate the at least one request; in response to validating the request, performing a merge operation that merges the one or more code changes with the second branch of software code; and deploying the second branch of software code to at least one cloud-based computing environment based at least in part on a version tag that identifies the merge operation corresponding to the request and a previous version tag that identifies a most recent merge operation completed for a previously deployed version of the second branch of software code.

Abstract: The present disclosure relates to a method for generating a program for use in an accelerator for deep learning. The method may include receiving, by a computing device, a deep learning application, generating an element-wise operation list included in the deep learning application, generating an intermediate expression from the element-wise operation list, and generating, based on the intermediate expression, a program for use in an accelerator for the deep learning application.

Abstract: A workflow tool to build, deploy and release application code of an application to any of one or more cloud computing platforms. The workflow tool is executed by one or more processors to generate a single configuration file for the application, the single configuration file comprising, using a common command-line interface (CLI) language: a build configuration defining a build tool used for building the application; a deploy configuration defining a cloud computing platform to which the application will be deployed; and a release configuration defining logic for releasing the application to one or more users of the cloud computing platform. The workflow tool further includes a release uniform resource locator (URL) generated by the first processor, the release URL providing access, by the cloud computing platform to the one or more users, to the application upon the releasing.

Abstract: A software update system includes: a vehicle data acquiring unit configured to acquire data on a state of a vehicle including data of a predetermined performance of the vehicle; an effect predicting unit configured to predict an effect on the predetermined performance which is obtained when update data of software used in the vehicle for improvement of the predetermined performance is applied to the vehicle based on the data acquired by the vehicle data acquiring unit; and a delivery unit configured to deliver the update data to the vehicle when the effect predicted by the effect predicting unit satisfies a predetermined reference and to withhold delivery of the update data to the vehicle when the effect predicted by the effect predicting unit does not satisfy the predetermined reference.

Abstract: A system and method for rule selection may include using one or more processors in a computer server, for a set of rules, each rule having one or more rule criteria, determining for each rule if the result of an interaction between interaction metadata and the one or more rule criteria includes one or more criteria. For each rule where the intersection includes one or more criteria, a complement result may be calculated between the result of the intersection and the rule criteria; and based on the complement result, wherein if the complement result is null, the rule may be selected.

Abstract: An application platform system and method. A data synchronization instance manages a reference state object for a data synchronization system (DSS) account. A first local state object is stored at a first application system. The first application system receives update notifications provided by the DSS. A second local state object is stored at a second application system. The second application system receives update notifications provided by the DSS. An application instruction of the first application system is transformed into a state update, and the first local state object is modified to include the state update. The state update is provided to the DSS via a local update notification. The reference state object is modified to include the state update. The state update is provided to the second application system via a reference update notification. The second local state object is modified to include the state update.

Abstract: A system can determine to uninstall a patch that is installed on a computer, wherein an image of an operating system of the computer and a package that was used to install the patch are cryptographically signed. The system can further identify a first stored location of the package, and a second stored location of the image of the operating system. The system can further extract a file manifest from the package at the first stored location. The system can further extract versions of files identified on the file manifest from the image of the operating system at the second stored location, resulting in extracted versions of files. The system can further store the extracted versions of files to a temporary storage location. The system can further move the extracted versions of files in the temporary storage location to respective original running locations of the versions of files.

Abstract: A computing device is disclosed herein. The computing device includes a memory that stores processor executable instructions for an application development platform and a context-aware undo-redo service of the application development platform. The computing device includes a processor that executes the processor executable instructions to cause the computing device to receive a first invocation of an undo operation with respect to environment variables on screens. The computing device further navigates, according to an active context, to a configuration screen of the screens to make the configuration screen visible in response to the first invocation. The configuration screen shows a portion of the environment variables. The computing device also receives a second invocation of the undo operation and executes the undo operation in response to the second invocation to reverse changes to the portion of the environment variables shown by the configuration screen while the configuration screen is visible.

Abstract: A computer that dynamically maps traffic to different versions of software is described. During operation, the computer may receive first traffic from a first tenant and second traffic from a second tenant, where the first traffic and the second traffic are addressed to a common computing environment that supports multiple, different tenants. Then, the computer may identify the first tenant based at least in part on first information included in the first traffic, and may identify the second tenant based at least in part on second information included in the second traffic. Moreover, the computer may dynamically map the first traffic to a first version of software and the second traffic to a second version of the software, where the dynamic mapping occurs at runtime. Note that the computer may concurrently execute the first version of the software and the second version of the software.

Abstract: A device of updating library required by a testing program for testing and a method thereof are disclosed. A system configuration file of an old testing device is modified to make the old testing device enter a testing mode when booting, so that library supporting the testing program of an application is updated in the testing mode; when the application is executed, the testing program is connected to the to-be-tested device for performing testing, so that the old testing device can be updated through network without dedicated hardware, and the technical effect of reducing the time and labor cost for updating the old testing device can be achieved.

Abstract: Systems and methods for scheduling and deploying patches to electronic device assets are provided. A patch management system identifies scheduling for patching and causes a patch scripting service to apply patching tags to the electronic device assets. A patch deployment service deploys patches based upon the applied tags.

Abstract: A device tracks accesses to pages of code executed by processors and modifies a portion of the code without terminating the execution of the code. The device is connected to the processors via a coherence interconnect and a local memory of the device stores the code pages. As a result, any requests to access cache lines of the code pages made by the processors will be placed on the coherence interconnect, and the device is able to track any cache-line accesses of the code pages by monitoring the coherence interconnect. In response to a request to read a cache line having a particular address, a modified code portion is returned in place of the code portion stored in the code pages.

Abstract: Systems, devices, and methods are provided for fingerprinting requests, such as transaction records. A transaction record or other suitable request may be parsed to identify a parameter values for a set of fields relevant to fingerprinting. A transaction record representation may be generated based on transaction input data, transaction output data, intermediate data, or combinations thereof. A fingerprint may be generated from the transaction record implementation. Fingerprints may be used to identify various test cases that can be used for regression testing.