Abstract: The disclosed computer-implemented method can include reaching, by a chiplet involved in carrying out an operation for a process, a synchronization barrier. The method can additionally include receiving, by the chiplet, dedicated control messages pushed to the chiplet by other chiplets involved in carrying out the operation for the process, wherein the dedicated control messages are pushed over a control network by the other chiplets. The method can also include advancing, by the chiplet, the synchronization barrier in response to receipt of the dedicated control messages. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: A number of cells of a control plane of a virtualized computing service are set up, including a first cell with one or more request processing nodes, a local instance of a data store, and metadata indicating a set of virtualization hosts. A request processer transmits a request for a virtual machine to the first cell. A request processor of the cell initiates a workflow to launch the virtual machine using a virtualization host; the workflow includes storing a record of the request in the local instance of the data store.

Abstract: A method and apparatus for processing a service using a request in a computer readable storage medium. An embodiment of the method comprises: acquiring a service using a request for a vehicle-mounted device; determining a to-be-processed request from unprocessed service using requests; determining, using a preset navigation status service layer, a service type to which the to-be-processed request belongs; and calling, using the navigation status service layer, a service module corresponding to the service type to process the to-be-processed request, the navigation status service layer recording corresponding relationships between different service types and different service modules.

Abstract: A set of configurable sync groupings (which may be referred to as sync zones) are defined. Any of the processors may belong to any of the sync zones. Each of the processor comprises a register indicating to which of the sync zones it belongs. If a processor does not belong to a sync zone, it continually asserts a sync request for that sync zone to the sync controller. If a processor does belong to a sync zone, it will only assert its sync request for that sync zone upon arriving at a synchronisation point for that sync zone indicated in its compiled code set.

Abstract: A method, computer program product, and computer system for implementing a backend service for blocking free processing of physical entities events, including add, remove, update, query. Physical entities blocking delays may be delegated to maintenance tasks, which may run under a single thread with a scheduler and may merge successive pending events.

Abstract: Methods, apparatus, and processor-readable storage media for generating temporal predictions for provisioning cloud resources using trained machine learning techniques are provided herein. An example computer-implemented method includes training one or more machine learning techniques using historical data pertaining to cloud resource provisioning and related order information; obtaining data pertaining to at least one order related to one or more hardware components, wherein the at least one order includes one or more cloud resources; generating at least one temporal prediction for provisioning the one or more cloud resources, in fulfillment of the at least one order, by processing at least a portion of the obtained data using the one or more trained machine learning techniques; and performing one or more automated actions based at least in part on the at least one generated temporal prediction.

Abstract: A system can maintain a first data center in a first physical location that comprises first compute hardware, and a second data center in a second physical location that comprises second compute hardware. The system can establish an overlay network that spans the first data center and the second data center. The system can establish a group of virtualized volume identifiers that spans the first data center and the second data center, and that virtualizes physical storage volumes. The system can determine whether to process a customer virtualized workload on the first data center or on the second data center to produce a selected location, wherein the customer virtualized workload is configured to be processed on the first data center and to be processed on the second data center. The system can process the customer virtualized workload at the selected location.

Abstract: A method for data sequence prediction and resource allocation includes determining, by a memory system, a plurality of resource parameters associated with operation of the memory system and determining respective time intervals associated with usage patterns corresponding to the memory system, the respective time intervals being associated with one or more sets of the plurality of resource parameters. The method further includes determining, using the plurality of resource parameters, one or more weights for hidden layers of a neural network for the respective time intervals associated with the usage patterns and allocating computing resources within the memory system for use in execution of workloads based on the determined one or more weights for hidden layers of the neural network.

Abstract: A request-response based paravirtualized I/O system and method relating to the fields of virtualization and cloud computing includes a request-response application, a front-end drive module and a back-end drive module. The front-end drive module and the back-end drive module interact by means of a transmit queue and a receive queue. The request-response application generates an I/O request, and the front-end drive module writes the I/O request into the transmit queue. The system has two operating modes: a notification mode and a polling mode. The system operates by default in the notification mode. When the request-response application issues a connection establishment or service request, the system switches to the polling mode. This system and method introduce an optimistic polling mechanism, combine the advantages of the notification mode and the polling mode, reduce the number of VM exits and decrease wasting of computing resources, thus improving data path performance.

Abstract: A virtual machine deployment method, a virtual machine live migration method, a VMM upgrading method, a server, and a computer-readable storage medium are disclosed. The virtual machine deployment method includes: establishing mapping between a host virtual address (HVA) space for a post-upgrading virtual machine and a host physical address (HPA) space for a pre-upgrading virtual machine according to a mapping relationship between a HVA space for the pre-upgrading virtual machine and the HPA space for the pre-upgrading virtual machine. The post-upgrading virtual machine is deployed on a post-upgrading virtual machine monitor (VMM), and the post-upgrading virtual machine is identical in memory configuration with the pre-upgrading virtual machine running on a pre-upgrading VMM (S110).

Abstract: Systems, methods, and devices are described for intelligent outage evaluation and insight management for monitoring and incident management systems. A set of incident logs associated with a cloud computing service are received by a cloud computing service provider system. Each incident log includes impact data corresponding to a respective incident that occurred during an evaluation period. The impact data includes an interruption count representative of a number of interruptions on resources caused by the respective incident. A total active resource time representative of resource activity during the evaluation period is determined. An interruption rate is determined for the evaluation period by summing interruption counts to generate a total interruption count and dividing the total interruption count by the total active resource time. An actionable insight applicable to the cloud computing service is generated based on the interruption rate and the incident logs.

Abstract: The present invention relates to a method for a processor and a computer system to enter system management mode (SMM). The method is applied to the processor, which includes at least one logical core. The method includes entering the SMM in response to a system management interrupt (SMI), storing current state information to a corresponding state save area; setting operation mode to target operating mode, and executing the SMI handler under the target operating mode. The address of the state save area, the address of the core configuration information memory space, and the address of the SMI handler can be determined directly. Each logical core may enter the target operating mode directly through hardware setting every time after entering the SMM, and thus does not need to perform a fixed operation of mode switching when executing the SMI handler, which may improve the execution efficiency of the SMI handler.

Abstract: A method and apparatus for scaling a Network Based Media Processing (NBMP) workflow on a cloud platform. The method includes obtaining an input workflow including input media stream, determining at least one of scaling type information or transition type information for the input work flow; generating a updated workflow based on the determined at least one of the scaling type information or the transition type information and processing the updated workflow.

Abstract: Methods and systems include processors and hardware accelerators. The processor initiates a first process in a first hardware accelerator configured to aid the processor in performing the first process. The processor initiates the first process using one or more interface registers. The processor performs additional processing while the first hardware accelerator performs the first process after initiation of the first process. The processor also initiates a second process in a second hardware accelerator configured to aid the processor in performing a second process. Moreover, the processor initiates the second process using the one or more interface registers.

Abstract: This application discloses a graph computing method and apparatus that support concurrent graph computing using a plurality of algorithms. A plurality of subgraphs of a graph are loaded into a plurality of computing units, and the plurality of computing units executes a plurality of algorithms in parallel, so that the same graph can be shared by the plurality of algorithms, and the plurality of algorithms can be executed in parallel on the same graph. In this way, the delay caused when one algorithm needs to executed after execution of another algorithm ends is avoided. The overall efficiency of performing graph computing is improved, and the overall time of performing graph computing is reduced.

Abstract: A method comprising a set of discrete rendering tasks having an initial task order to be executed by a graphics processing unit, a GPU, detecting a dependency between a dependency group of two or more rendering tasks, a dependency being that one or more rendering tasks of the dependency group earlier in the initial task order depend upon an execution result of one or more rendering tasks of the dependency group later in the initial task order; ordering the set of rendering tasks, in the ordered set of rendering tasks, a dependency group of rendering tasks is ordered such that rendering tasks of a dependency group which depend upon one or more other tasks of the dependency group are ordered after the one or more other tasks upon which those rendering tasks depend; and issuing the ordered set of rendering tasks to the GPU for execution.

Abstract: A computer-implemented method includes receiving a neural network model that includes memory load operations and a plurality of computation operations, selecting a memory load operation having an arithmetic intensity factor (AIF) greater than a threshold value from the memory load operations, grouping computation operations associated with data loaded by the selected memory load operation into two or more clusters of computation operations, and incorporating an instance of the selected memory load operation before each cluster of the two or more clusters of computation operations in the neural network model.

Abstract: Unified automation of application development and delivery is provided. An automation pipeline execution coordinator may define a pipeline specification that includes actions to be performed, a triggering event definition and specification for determining execution context. The coordinator may concurrently detect triggering events for multiple pipelines matching the pipeline specification, and responsive to the detecting, determine execution contexts for the pipelines. The coordinator may then execute the multiple pipelines, where execution may proceed independently for pipelines with differing execution contexts. For pipelines sharing an execution context, execution of various actions of the respective pipelines may be coordinated. Execution context may be determined according to the specification for determining execution context, which may include an overridable default specification that determines context by locations of source data related to the triggering event.

Abstract: Methods and electronic devices for batching requests for processing by a processing unit are disclosed. The method includes acquiring a queue of requests to be processed and executing a current iteration for a given batch of requests including the given request from the queue. During the current iteration, the method includes determining a first cost value of adding a first candidate request from the queue to the given batch, determining a second cost value of adding a second candidate request from the queue to the given batch, and selectively adding a target request amongst the first candidate request and the second candidate request to the given batch. The target request being associated with a lowest one of the first cost value and the second cost value.

Abstract: Transactional Lock Elision allows hardware transactions to execute unmodified critical sections protected by the same lock concurrently, by subscribing to the lock and verifying that it is available before committing the transaction. A “lazy subscription” optimization, which delays lock subscription, can potentially cause behavior that cannot occur when the critical sections are executed under the lock. Hardware extensions may provide mechanisms to ensure that lazy subscriptions are safe (e.g., that they result in correct behavior). Prior to executing a critical section transactionally, its lock and subscription code may be identified (e.g., by writing their locations to special registers). Prior to committing the transaction, the thread executing the critical section may verify that the correct lock was correctly subscribed to. If not, or if locations identified by the special registers have been modified, the transaction may be aborted.