Abstract: Methods, apparatus, computer program products for resource scheduling are provided. The method comprises: receiving a workload request; publishing the information of the workload to a workload billboard accessible to a plurality of computer hosts, each of the plurality of computer hosts being associated with a corresponding proxy configured to manage the resource scheduling of the computer host; receiving a request to schedule at least a portion of the workload from a proxy; and sending the portion of the workload to the computer host associated with the proxy.

Abstract: Provided is a hybrid near-memory processing system including a GPU, a PIM-HBM, a CPU, and a main memory. An embedding vector is loaded through the GPU and the PIM-HBM, an embedding table is divided and stored in the main memory and the HBM in a training process for inference of a recommendation system, an embedding lookup operation is performed in the main memory or the HBM according to a location of a necessary embedding vector in an inference process of the recommendation system, an additional embedding manipulation operation is performed in the CPU and the PIM with respect to the embedding vector of which the embedding lookup operation is completed, embedding vectors processed through embedding manipulation are finally concatenated in the PIM to generate an embedding result, and the embedding result is transmitted to the GPU to derive a final inference result through a top multiplayer perceptron (MLP) process.

Abstract: Described are techniques for scheduling tasks on a heterogeneous system on a chip (SoC). The techniques including receiving a directed acyclic graph at a meta pre-processor associated with a heterogeneous system-on-chip and communicatively coupled to a scheduler, where the directed acyclic graph corresponds to a control flow graph of respective tasks associated with an application executed by the heterogeneous system-on-chip. The techniques further including determining, using a learning agent implementing machine learning algorithms, a rank for a respective task in the directed acyclic graph, wherein the learning agent receives as input the directed acyclic graph, constraints associated with the directed acyclic graph, and heuristics regarding previously completed tasks. The techniques further including providing the respective task to the scheduler for execution on the heterogeneous system-on-chip according to the rank.

Abstract: A computer-implemented method and system for resilient adaptive biased locking. The method includes adding, in a system including an adaptive lock reservation scheme having a learning state, a component comprising a per class counter that counts, collectively, a number of learning failures and a number of revocation failures. An embodiment includes initializing the per class counter upon loading a class with a predetermined value representing at least one of a maximum number of learning failures and cancellation instances associated with the class. An embodiment includes initializing, based on a determination of an operational state of the per class counter for an object transitioning from one of the learning state and a biased state to a flatlock state, a lock word of the object directly to the flatlock state while bypassing the biased state.

Abstract: A processor includes a task scheduling unit and a compute unit coupled to the task scheduling unit. The task scheduling unit performs a task dependency assessment of a task dependency graph and task data requirements that correspond to each task of the plurality of tasks. Based on the task dependency assessment, the task scheduling unit schedules a first task of the plurality of tasks and a second proxy object of a plurality of proxy objects specified by the task data requirements such that a memory transfer of the second proxy object of the plurality of proxy objects occurs while the first task is being executed.

Abstract: A performance influence involved in system transition is evaluated in consideration of a timer set for each processing section. An evaluation device 1 includes a storage device 10 that stores processing section data 11 in which a maximum time from start to expiration of a timer is associated with an identifier of a processing section in which the timer is set, an average waiting time calculating unit 21 that calculates an average waiting time of a service request based on a turnaround time necessary for processing in the accumulation device 4 for each processing section, and as evaluation unit 22 that evaluates that data used in the processing section is not separable to the accumulation device 4 when the maximum time of the timer set in the processing section is less than the sum of the average waiting time and a traffic amount per unit time.

Abstract: The disclosed system and methods are directed to a data-driven choreography of microservices sagas that enable the composition of microservices without depriving each individual microservice of its independence.

Abstract: An electronic apparatus includes: a memory; a storage; and a processor, wherein: the electronic apparatus is configured to execute a plurality of processes as data of the plurality of processes is loaded into the memory based on execution of at least one program stored in the storage, the processor is configured to: identify a function currently running among a plurality of functions providable by the electronic apparatus, and based on a relationship between the plurality of processes and the identified function, terminate at least one process among the plurality of running processes, and allow a storage area of the memory loaded with the data of the terminated process to be available for another process.

Abstract: A resource processing method and apparatus are proposed in the embodiment of the present disclosure. The method includes: obtaining a resource allocation scenario; querying a corresponding strategy combination configuration according to the resource allocation scenario; executing each allocation strategy to determine a target resource pool allocated with the allocation strategy; obtaining resources corresponding to each allocation strategy from the target resource pool allocated with the allocation strategy; executing the combination operation on the resources corresponding to the allocation strategy to obtain target resources.

Abstract: The system of the present technology includes an embodiment that provides a host audio, video and control operating system configured to establish or interact with one or more virtual machines, each with a guest operating system.

Abstract: An electronic device includes a memory; a plurality of clients; at least one arbiter circuit; and a management circuit. A given client of the plurality of clients communicates a request to the management circuit requesting an allocation of memory access bandwidth for accesses of the memory by the given client. The management circuit then determines, based on the request, a set of memory access bandwidths including a respective memory access bandwidth for each of the given client and other clients of the plurality of clients that are allocated memory access bandwidth. The management circuit next configures the at least one arbiter circuit to use respective memory access bandwidths from the set of memory access bandwidths for the given client and the other clients for subsequent accesses of the memory.

Abstract: A method for graphics processing, wherein a graphics processing unit (GPU) resource is allocated among applications, such that each application is allocated a set of time slices. Commands of draw calls are loaded to rendering command buffers in order to render an image frame for a first application. The commands are processed by the GPU resource within a first time slice allocated to the first application. The method including determining at least one command has not been executed at an end of the first time slice. The method including halting execution of commands, wherein remaining one or more commands are not processed in the first time slice. A GPU configuration is preserved for the commands after processing a last executed command, the GPU configuration used when processing in a second time slice the remaining commands.

Abstract: The present approach relates to systems and methods for facilitating run time predictions for cloud-computing automated tasks (e.g., automated tasks), and using the predicted run time to schedule resource locking. A predictive model may predict the automated task run time based on historical run time to completion, and the run time may be updated using machine learning. Resource lock schedules may be determined for a queue of automated tasks utilizing the resource based on the predicted run time for the various types of automated tasks. The predicted run time may be used to reserve a resource for the given duration, such that the resource is not available for use for another task.

Abstract: Approaches in accordance with various embodiments can reduce scheduling delays due to concurrent processing requests, as may involve VSyncs in multi-streaming systems. The software synchronization signals can be staggered relative to each other by offsetting an initial synchronization signal. These software synchronization signals can be readjusted over time such that each synchronization signal maintains the same relative offset, as may be with respect to other applications or containers.

Abstract: Described herein are systems, methods, and software to enhance the management of virtual nodes in a computing environment. In one example, a system is configured to mount storage repositories to a host. Once mounted the system may identify a request to initiate a virtual cluster on the host and may identify permissions associated with the virtual cluster. The system further determines at least one storage repository from the storage repositories for the virtual cluster based on the permissions and initiate execution of the virtual cluster with access to at least one file system corresponding to the at least one storage repository.

Abstract: An electronic device for executing various operating systems is provided. The electronic device includes first and second hardware devices, a first operating system (OS), a second OS different from the first OS, and a processor configured to control the first hardware device to process first data from a first program executed on the first OS, obtain a command for executing the second OS, generate a container for executing the second OS based on a kernel of the first OS in response to the command for executing the second OS, execute the second OS on the generated container, execute a second program on the second OS, obtain second data regarding the second program from the second OS via socket communication by a control application installed on the first OS, and control the second hardware device to process the second data regarding the second program based on the first OS using the installed control application.

Abstract: An apparatus includes a memory and one or more processors. The one or more processors are configured to receive a first computing task for execution, and, during execution of the first computing task (i) to monitor Operating System (OS) calls issued by the first computing task, (ii) to identify, based on the monitored OS calls, one or more outputs produced by the first computing task, and (iii) and to store the one or more outputs of the first computing task in the memory. The one or more processors are further configured to receive a second computing task for execution, and, upon ascertaining that the second computing task will produce at least one output that is identical to a stored output of the first computing task, to reuse the stored output of the first computing task as an output or the second computing task.

Abstract: An information processing system includes: a storage device configured to store first clock time scheduled for execution of a task; and a processing device configured to: execute a task at second clock time earlier than first clock time scheduled for execution of the task; control the executing of the task not to execute the task at the first clock time when data used in the task is not updated in a period from the second clock time to the first clock time; and control the executing of the task to re-execute the task either at the first clock time or at third clock time earlier than the first clock time when the data is updated in the period from the second clock time to the first clock time.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for efficiently using computing resources when responding to content requests. Methods include using a prioritization model and a specified threshold specifying the maximum allowable negative outcome for a content provider, to determine whether a received content request is a low priority request. Methods further include throttling access to computing resources to respond to low priority requests, while providing access to computing resources for other content requests. Methods also include regularly updating the prioritization model and the specified threshold based on data for a new set of content requests.

Abstract: Provided herein are an electronic apparatus and a controlling method thereof. An electronic apparatus according to the disclosure includes a communicator, a memory storing information on a recipe wherein a plurality of unit functions for provision of a service are combined, and a processor configured to, based on receiving information for a unit function that can be performed at each electronic apparatus from each of a plurality of electronic apparatuses through the communicator, identify a plurality of electronic apparatuses matched to the plurality of unit functions included in the recipe based on the received information, and control the communicator to transmit a signal for performing each matched unit function to each of the plurality of identified electronic apparatuses.