Abstract: A computer-implemented method, a computer program product, and a computer processing system are provided for selecting from among multiple Graphics Processing Unit (GPU) execution modes for a Neural Network (NN) having a size greater than a threshold size. The multiple GPU execution modes include a normal memory mode, an Out-of-Core (OoC) execution mode, and a Unified Memory (UM) mode. The method includes starting an execution on the NN with the UM mode and measuring the memory usage for each of layers of the NN. The method further includes selecting an execution mode based on the memory usage of all of the layers.

Abstract: One or more processors receive resource type and capability information and activity information of workloads of a domain. A first model is generated and trained to map the resource information to the activity information of domain workloads. The activity information is decomposed into a set of activity core elements (ACEs). The one or more processors generate a second model, wherein the second model is trained to predict a set of resource types and resource capabilities of the respective resource types, based on an input of the first set of ACEs decomposed from the activity information of the workloads of the domain. The one or more processors receive a second set of ACEs that are decomposed from activities associated with an unprecedented workload, and the one or more processors generate a predicted set of resources to perform the second set of ACEs.

Abstract: A memory device includes a plurality of memory dies, each memory die of the plurality of memory dies comprising a memory array and control logic. The control logic comprises a plurality of processing threads to execute memory access operations on the memory array concurrently, a thread selection component to identify one or more processing threads of the plurality of processing threads for a power management cycle of the associated memory die and a power management component to determine an amount of power associated with the one or more processing threads and request the amount of power during the power management cycle.

Abstract: Various examples are disclosed for performing cross-cluster load balancing. In some aspects, a workload is selected for cross-cluster migration. A destination cluster is identified for a migration of the workload from a source cluster to the destination cluster. A cross-cluster migration recommendation is generated to migrate the workload from the source cluster to the destination cluster.

Abstract: A system including a distributed network of one or more virtual machines, having a first portion of committed virtual machines dedicated to a user and a second portion of on-demand virtual machines. The system may also include a workload scheduler configured to receive workloads associated with the user. The scheduler may determine whether to schedule a given workload to be performed by a combination of virtual machines in the first and second portions, or by virtual machines included in only the first portion. The given workload may be scheduled to be performed by virtual machines in only the first portion at a first time if a sum of an anticipated resource consumption level of the given workload and a first consumption level of the first portion of virtual machines at the first time is less than or equals a total amount of resources included in the first portion.

Abstract: A method may include receiving, by a privileged component executed by a processing device, bytecode of a packet processing component from an unprivileged component executed by the processing device, analyzing, by the privileged component, the bytecode of the packet processing component to identify whether the bytecode comprises a first command that returns a redirect, analyzing, by the privileged component, the bytecode of the packet processing component to identify whether the bytecode comprises a second command that returns a runtime computed value, and responsive to determining that the bytecode comprises the first command or the second command, setting a redirect flag maintained by the privileged component.

Abstract: Embodiments of the present disclosure disclose a method and apparatus for acquiring information. The method may include: acquiring cloud resource use information; comparing the cloud resource use information with a use rate threshold value, to obtain a comparison result; obtaining use state information of a cloud resource corresponding to the cloud resource use information according to the comparison result; and generating cloud resource state information based on the use state information.

Abstract: A computer-implemented method, a computer program product, and a computer system for placements of workloads in a system of multiple platforms as a service. A computer detects a triggering event for modifying a matrix that pairs respective workloads on respective platforms and includes attributes of running respective workloads on respective platforms. The computer recalculates the attributes in the matrix, in response to the triggering event being detected. The computer determines optimal placements of the respective workloads on the respective platforms, based on information in the matrix. The computer places the respective workloads on the respective platforms, based on the optimal placements.

Abstract: Techniques for transaction tracking for a high availability architecture are described herein. An aspect includes receiving a first request from a client, the first request corresponding to a start of a transaction having transaction affinity. Another aspect includes, based on receiving the first request from the client, generating a transaction tracking token. Another aspect includes sending the first request with the transaction tracking token to a gateway. Another aspect includes receiving a first response corresponding to the first request from the gateway, the first response including middleware instance information corresponding to a middleware instance, wherein a plurality of subsequent requests from the client corresponding to the transaction are processed by the middleware instance corresponding to the middleware instance information.

Abstract: A computer-implemented method, a computer program product, and a computer system for optimizing workload placements in a system of multiple platforms as a service. A computer first places respective workloads on respective platforms that yield lowest costs for the respective workloads. The computer determines whether mandatory constraints are satisfied. The computer checks best effort constraints, in response to the mandatory constraints being satisfied. The computer determines a set of workloads for which the best effort constraints are not satisfied and determines a set of candidate platforms that yield the lowest costs and enable the best effort constraints to be satisfied. From the set of workloads, the computer selects a workload that has a lowest upgraded cost and updates the workload by setting an upgraded platform index.

Abstract: Method, device, and computer program product for managing processes are disclosed. The method may include, in response to detecting a job to be executed in a job processing system, determining attribute information related to execution of a message queue in the job processing system; and determining, based on the attribute information, allocation information for a process to execute the job. The number of processes can be dynamically adjusted according to the current processing capacity of the job processing system, thereby improving the use of the resources of the job processing system and meanwhile avoiding crashing of the job processing system.

Abstract: This disclosure describes methods, devices, systems, and procedures in a computing system for capturing a configuration state of an operating system executing on a central processing unit (CPU), and offloading memory management tasks, based on the configuration state, to a resource management unit such as a system-on-a-chip (SoC). The resource management unit identifies a status of a resource requiring memory swapping based on the captured configuration state of the operating system. The resource management unit then swaps the memory to alleviate the CPU from processing the swap thereby improving overall computing system performance.

Abstract: In a computing system, an application thread is executed on a hardware thread. Based on a configuration of the computing system, a first threshold is determined comprising a threshold percentage of execution time spent servicing a set of interrupts to the application thread relative to a total execution time for the hardware thread. For the hardware thread, a length of a first time period spent servicing an interrupt in the set of interrupts and a length of a second time period spent executing the application thread are measured. A cumulative percentage of execution time spent in the first time period relative to execution time spent in the first time period and the second time period is calculated. Responsive to the cumulative percentage being above the threshold percentage, interrupt servicing on the hardware thread is disabled.

Abstract: System, methods, and other embodiments described herein relate to improving scheduling of computing tasks in a mobile environment for a vehicle. In one embodiment, a method includes receiving an offloading request associated with a computing task from the vehicle, wherein the offloading request includes context information and a task descriptor related to the computing task. The method also includes scheduling the computing task to execute on a server if the context information and the task descriptor satisfy criteria for using computing resources associated with the server for the vehicle. The method also includes partitioning the computing task into subtasks if the context information satisfies the criteria. A machine learning module may decide partitions of the computing task according to the context information.

Abstract: Systems, methods, devices, and other techniques for managing a computing resource shared by a set of online entities. A system can receive a request from a first online entity to reserve capacity of the computing resource. The system determines a relative priority of the first online entity and identifies a reservation zone that corresponds to the relative priority of the first online entity. The system determines whether to satisfy the request based on comparing (i) an amount of the requested capacity of the computing resource and (ii) an amount of the portion of unused capacity of the computing resource designated by the reservation zone that online entities having relative priorities at or below the relative priority of the first online entity are permitted to reserve.

Abstract: Zero copy message reception for devices is disclosed. For example, a host has a memory, a processor, a supervisor, and a device with access to device memory addresses mapped in a device page table via an IOMMU. An application has access to application memory addresses and is configured to identify a first page of memory addressed by an application memory address to share with the device as a receiving buffer to store data received by the device for the application, where the first page is mapped to a first device memory address in a first device page table entry (PTE). A supervisor is configured to detect that the first application has disconnected from the device, and in response to detecting the application disconnecting, to update the first device PTE to address a second page instead of the first page.

Abstract: Systems and methods for performing parallel computation are disclosed. The system can include: a task manager; and a plurality of cores coupled with the task manager and configured to respectively perform a set of parallel computation tasks based on instructions from the task manager, wherein each of the plurality of cores further comprises: a processing unit configured to generate a first output feature map corresponding to a first computation task among the set of parallel computation tasks; an interface configured to receive one or more instructions from the task manager to collect external output feature maps corresponding to the set of parallel computation tasks from other cores of the plurality of cores; a reduction unit configured to generate a reduced feature map based on the first output feature map and received external output feature maps.

Abstract: A request manager analyzes API calls from a client to a host application for state and performance information. If current utilization of host application processing or memory footprint resources exceed predetermined levels, then the incoming API call is not forwarded to the application. If current utilization of the host application processing and memory resources do not exceed the predetermined levels, then the request manager quantifies the processing or memory resources required to report the requested information and determines whether projected utilization of the host application processing or memory resources inclusive of the resources required to report the requested information exceed predetermined levels. If the predetermined levels are not exceeded, then the request manager forwards the API call to the application for processing.

Abstract: Embodiments of systems and methods for managing performance optimization of applications executed by an Information Handling System (IHS) are described. In an illustrative, non-limiting embodiment, a method may include: identifying, by an IHS, a first application; assigning a first score to the first application based upon: (i) a user's presence state, (ii) a foreground or background application state, (iii) a power adaptor state, and (iv) a hardware utilization state, detected during execution of the first application; identifying, by the IHS, a second application; assigning a second score to the second application based upon: (i) another user's presence state, (ii) another foreground or background application state, (iii) another power adaptor state, and (iv) another hardware utilization state, detected during execution of the second application; and prioritizing performance optimization of the first application over the second application in response to the first score being greater than the second score.

Abstract: Disclosed herein are system, method, and computer program product embodiments for configuring a dynamic reassignment of an application flow across different computation layers based on various conditions. An embodiment operates by assigning a first rule of an application flow to a first computation layer of a plurality of computation layers. The embodiment assigns a second rule of the application flow to a second computation layer of the plurality of computation layers. The embodiment assigns a transition rule of the application flow to the first computation layer. The transition rule includes an action that causes the first rule of the application flow to be executed in the second computation layer of the plurality of computation layers based on a condition. The embodiment then transmits the application flow to the plurality of computation layers thereby causing the application flow to be configured for execution.