Methods, systems, and apparatus, including computer programs encoded on computer storage media, for scheduling operations represented as a computational graph on a distributed computing network. A method includes: receiving data representing operations to be executed in order to perform a job on a plurality of hardware accelerators of a plurality of different accelerator types; generating, for the job and from at least the data representing the operations, features that represent a predicted performance for the job on hardware accelerators of the plurality of different accelerator types; generating, from the features, a respective predicted performance metric for the job for each of the plurality of different accelerator types according to a performance objective function; and providing, to a scheduling system, one or more recommendations for scheduling the job on one or more recommended types of hardware accelerators.

Provided are a request processing unit to send a first process distribution request to other information processing apparatuses via a communication I/F unit, the first process distribution request being a request for ordering execution of a first process on a first; and an order destination selecting unit to receive, via the communication I/F unit, as a response to the first process distribution request, a first estimated reply time calculated as the time required to receive the transfer of the first data and to reply a first process result obtained by executing the first process on the first data, to use the first estimated reply time to select an order destination to which the execution of the first process is to be ordered by the plurality of information processing apparatuses, and to order the execution of the first process on the first data to the order destination via the communication interface unit.

Application lifecycle management based on real-time resource usage. A first plurality of resource values that quantify real-time computing resources used by a first instance of an application is determined at a first point in time. Based on the first plurality of resource values, one or more utilization values are stored in a profile that corresponds to the application. Subsequent to storing the one or more utilization values in the profile, it is determined that a second instance of the application is to be initiated. The profile is accessed, and the second instance of the application is caused to be initiated on a first computing device utilizing the one or more utilization values identified in the profile.

This disclosure relates to method and system for secure scheduling of workflows and virtual machine utilization in cloud. Scheduling of tasks in workflow comprises of heterogeneous and interdependent computational tasks. The method receives a set of workflows comprising of one or more heterogeneous tasks. Further, a set of parameters are extracted from each heterogeneous task to select a set of optimal VM type combination parameters and a set of security level combination parameters. The method selects the optimized combination of VM types, security service levels and task order. Further, a workflow schedule is generated for the tasks of the selected VM type combinations. The method further performs optimal selection of VM types and security services, with efficient schedule generation, and effectively reuses VM with reduced overall cost without delay in make span. Additionally, the method enhances security model with accurate risk estimation.

Techniques described herein relate to a method for performing computational offloads for composed information handling systems. The method includes obtaining, by a system control processor associated with a composed information handling system, a computational offload request associated with a dataset from an application executing on an at least one compute resource set; in response to obtaining the computational offload request: identifying a dataset location associated with the dataset in the composed information handling system; identifying resources of the composed information handling system capable of performing the computational offload request; selecting a resource of the resources to perform the computational offload; and initiating performance of the computational offload request on the selected resource.

The present disclosure relates to a method, device and computer program product for executing a job in an application system. Here, the application system comprises a first processing device and a second processing device, and a first response speed of the first processing device being lower than a second response speed of the second processing device. In a method, a job request is received from a user of the application system, the job request specifying that the job is to be executed in the application system; a job type of the job is determined, the job type describing a requirement of the user on a response speed for executing the job; a target processing device is selected from the first processing device and the second processing device in accordance with determining that the job type relates to a high response speed; and the job is assigned to the selected target processing device, so that the job is executed by the target processing device. By means of the above method, a processing device for processing a job is selected based on the type of the job, and further processing devices in the application system may be dispatched more effectively. Furthermore, there is provided a corresponding device and computer program product.

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for scheduling operations represented as a computational graph on a distributed computing network. A method includes: receiving data representing operations to be executed in order to perform a job on a plurality of hardware accelerators of a plurality of different accelerator types; generating, for the job and from at least the data representing the operations, features that represent a predicted performance for the job on hardware accelerators of the plurality of different accelerator types; generating, from the features, a respective predicted performance metric for the job for each of the plurality of different accelerator types according to a performance objective function; and providing, to a scheduling system, one or more recommendations for scheduling the job on one or more recommended types of hardware accelerators.

Disclosed is an SLO-aware artificial intelligence inference scheduler technology in a heterogeneous processor-based edge system. A scheduling method for a machine learning (ML) inference task, which is performed by a scheduling system, may include receiving inference task requests of multiple ML models with respect to an edge system composed of heterogeneous processors and operating heterogeneous processor resources of the edge system based on a service-level objective (SLO)-aware-based scheduling policy in response to the received inference task requests.

Embodiments of the present invention provide a system for leveraging distributed register technology to securely monitoring, tracking, and recommending utilization of resources. The system is configured for gathering one or more input parameters from one or more entity systems, collecting activity data from one or more third party systems, analyzing the activity data collected from the one or third party systems, generating one or more recommendations based on the one or more input parameters and analyzing the activity data, wherein the one or more recommendations are associated with one or more activities, estimating resource usage for the one or more recommendations, and allocating resources to the one or more recommendations.

Multi-cloud deployment strategy is based on automated analysis of context and requirements for an activity workload. The activity workload is defined by user input including information regarding project cost, performance requirements, and geographical preferences. Selection of cloud-based resources for handling the activity workload is based in part on service availability record, projected cost of resources, and physical geographic locations. A cloud services registry provides cloud service provider data for selection to perform aspects of the activity workload.