A non-transitory computer-readable recording medium stores a management program for causing a computer to execute a process the includes receiving designation of a first resource used by a specific server included in a plurality of servers, calculating a first estimated value of power to be consumed by the plurality of servers when the specific server executes first processing using the first resource, determining whether the first estimated value exceeds a first upper limit value of power to be consumed by the plurality of servers, setting in the specific server, when determined that the first estimated value does not exceed the first upper limit value, a second estimated value of power to be consumed by the specific server when the first processing is executed using the first resource, as a second upper limit value of power to be consumed by the specific server.

Determining utilization of a computing component, including: determining, of the power limits of the respective registers, a lowest power limit; determining an instantaneous power use of the computing component; determining a power-based utilization of the computing component based on i) the instantaneous power use of the computing component and ii) the lowest power limit indicated by the registers; identifying an instantaneous temperature and temperature limit of the computing component, and an ambient temperature of an environment of the computing component; determining a temperature-based utilization of the computing component based on the instantaneous temperature and the temperature limit of the computing component, and the ambient temperature of the environment; determining the utilization of the computing component based on a greater of the power-based utilization of the computing component and the temperature-based utilization of the computing component, and in response, adjusting execution of computer-implemented actions performed at the computing component.

Within a processing system, thread count asymmetries manifest when one or more cores of a processing device are disabled. To determine such thread count asymmetries, discovery operations are performed to determine thread count asymmetries for one or more hierarchy levels of a processing device based on a number of threads per enumerated instance within the hierarchy level. In response to the determining a thread count asymmetry, one thread identifier for each enumerated instance within the asymmetric hierarchy level is defined to determine a representation of the asymmetry. Using the representation of the symmetry, software tasks associated with one or more application within the processing system are performed.

A vehicle-based apparatus for noise injection and monitoring includes a first processing device coupled to a plurality of components and a second processing device coupled to the plurality of components. The second processing device injects noise into at least one component among the plurality of components and monitors behavior of the at least one component in response to the injected noise. The second processing device determines, based, at least in part, on the monitored behavior of the at least one component responsive to the injected noise, a susceptibility to the injected noise exhibited by the at least one component and transmits signaling indicative of the determined susceptibility to the injected noise exhibited by the at least one component to the first processing device.

A method by a computing system associated with a set of disjoint devices that includes at least one wearable device includes receiving a request to perform a task. The method further includes determining, based on sensor data associated with the set of disjoint devices, a thermal-constraint differential for each device of the set of disjoint devices. The method further includes determining a plurality of workload assignments needed to be performed to accomplish the task. The method further includes distributing, based on the thermal-constraint differentials of the set of disjoint devices, the plurality of workload assignments to one or more devices of the set of disjoint devices to satisfy one or more power or thermal constraints associated with each device of the set of disjoint devices. The method further includes performing the task by causing the one or more devices to execute the distributed plurality of work assignments.

A computing method and device for large-scale computing is provided. The computing device includes at least one processing device configured to perform an operation related to a neural network, a sensor configured to sense an electrical characteristic of the at least one processing device, an operating frequency of the at least one processing device, and a temperature of the at least one processing device, and a processor configured to calculate a workload to be allocated to the at least one processing device based on an operating mode of the at least one processing device, the electrical characteristic of the at least one processing device, the operating frequency of the at least one processing device, and the temperature of the at least one processing device, and control the electrical characteristic, the operating frequency, and the temperature based on the operating mode and the workload.

A processing system adjusts an operating state of one or more processors during execution of a workload based on a command paired with the workload. The command specifies a desired operating state or a performance or power efficiency target operational goal and is enqueued asynchronously with the workload. A power management controller reads the command synchronously with dispatching the workload to the processor. By asynchronously enqueuing the tag with the workload, the processing system tunes the operating state of the processor to reach higher performance, higher performance per watt, and/or higher energy efficiency during processing of the workload.

An apparatus comprises a processing device configured to select prioritization criteria for a plurality of virtual computing resources and to determine, for at least one criterion in the selected prioritization criteria, at least one directional correlation between the at least one criterion and at least one other criterion in the selected prioritization criteria. The processing device is also configured to generate a prioritization of the plurality of virtual computing resources utilizing a multi-criteria decision analysis algorithm. The multi-criteria decision analysis algorithm is based at least in part on the determined at least one directional correlation. The processing device is further configured to perform one or more scheduling operations for the plurality of virtual computing resources based at least in part on the generated prioritization of the plurality of virtual computing resources.

Systems include one or more critical datacenter connected to behind-the-meter flexible datacenters. The critical datacenter is powered by grid power and not necessarily collocated with the flexible datacenters, which are powered “behind the meter.” When a computational operation to be performed at the critical datacenter is identified and determined that it can be performed more efficiently or advantageously at a flexible datacenter, the computational operation is instead obtained by the flexible datacenters for performance. The critical datacenter and flexible datacenters preferably share a dedicated communication pathway to enable high-bandwidth, low-latency, secure data transmissions. In some situations, a computational operation is supported by multiple datacenters in a redundant arrangement, such as multiple flexible datacenters.

Provided are an apparatus and a method for controlling an interrupt rate for a processor based on processor utilization. Accordingly, it is possible to improve an I/O response latency and improve energy efficiency.