In one embodiment, an apparatus comprises a communication interface and a processor. The communication interface is to communicate with a plurality of devices. The processor is to: receive compressed data from a first device, wherein the compressed data is associated with visual data captured by sensor(s); perform a current stage of processing on the compressed data using a current CNN, wherein the current stage of processing corresponds to one of a plurality of processing stages associated with the visual data, and wherein the current CNN corresponds to one of a plurality of CNNs associated with the plurality of processing stages; obtain an output associated with the current stage of processing; determine, based on the output, whether processing associated with the visual data is complete; if the processing is complete, output a result associated with the visual data; if the processing is incomplete, transmit the compressed data to a second device.

A system and method for allocating software resources. Multiple tasks are received from a network in which each task requires at least one software resource. Each task is analyzed to determine the type of resource(s) required to execute each such task. The availability of the software resource(s) is determined and, if available, allocated to the requesting task. If the software resource(s) is not available, the task is stored in a queue until the software resource(s) becomes available.

Apparatuses, systems, and methods related to sharing a memory resource among physically remote entities are described. A system sharing a memory resource among physically remote entities may enable performance of functions, including automated functions critical for prevention of damage to a product, personnel safety, and/or reliable operation, based on increased access to data that may improve performance of a mission profile. For instance, one apparatus described herein includes a first vehicle configured to determine an availability of processing resources or memory capacity, or both, at the first vehicle based at least in part on a current operating mode of the first vehicle, receive a request from a second vehicle to use at least a portion of the processing resources or the memory capacity, or both, to perform a processing operation at a second vehicle, wherein the request from the second vehicle is associated with insufficient processing capability or memory capacity, or both, at the second vehicle, and perform at least a portion of the processing operation or allow access to the available memory capacity, or both, at the first vehicle in response to the request and based at least in part on determining the availability of the processing resources or the memory capacity, or both.

A computer-implemented method for scheduling a series of recurring events including: receiving one or more requests to allocate resource(s) to a series of recurring events, wherein the one or more requests specify, for each event, a corresponding desired time period over which the resource(s) are to be allocated, and the one or more requests further specify one or more adjustment criteria for defining, for one or more of the events, one or more permissibly adjusted time periods from the desired time period; obtaining, for each event, resource availability data indicative of an availability of the resource(s) during the desired time period; and, for each event: determining, based on the resource availability data, a viable time period, wherein the viable time period is either the desired time period or a permissibly adjusted time period that satisfies the one or more adjustment criteria; and allocating the resource(s) to the viable time period.

A system and a method for routing a message to an application over a connection oriented session in a Kafka messaging platform environment are provided. The method includes: acquiring a plurality of partitions from the Kafka messaging platform; designating a first partition from among the plurality of partitions as a sticky partition; generating a plurality of routing keys that are configured to route to the sticky partition; receiving a subscription from a service that corresponds to a first application; transmitting, to the first application, a first routing key that identifies the subscription from among the plurality of routing keys; and receiving messages from Kafka services that are routed by the first routing key to the first application. For any particular application or set of applications, a plurality of connection oriented sessions may be used to achieve load balancing and high availability.

Systems and methods for utilizing network hints to configure the operation of modern workspaces are described. In an embodiment, an Information Handling System (IHS) may include a processor and a memory coupled to the processor, the memory having program instructions stored thereon that, upon execution, cause the IHS to: receive, by a network performance service, a network configuration policy; determine, by the network performance service, one or more characteristics of network traffic generated by a selected one of a plurality of workspaces instantiated via a local management agent; receive, by the network performance service, one or more traffic hints; and execute a responsive action prescribed in the network configuration policy based, at least in part, upon: (i) the one or more characteristics of network traffic, and (ii) the one or more traffic hints.

A virtual resource management device (1) includes a scaling request reception unit (11) that receives a scale-in request; a virtual resource identification unit (12) that identifies a virtual resource to be deleted by using a predetermined virtual resource identification logic according to a deletion strategy indicated by the scale-in request; and a scaling processing unit (13) that transmits a scale-in preparation request including information on the identified virtual resource to the VNF to stop communication to the virtual resource, and in response to receiving a scale-in preparation completion response indicating the communication has been stopped from the VNF, deletes the virtual resource to be deleted.

An inference system applies a machine-learning transformer model to a batch of requests with variable input length or variable target length or variable internal sate length by selectively batching a subset of operations in the transformer model but processing requests in the batch individually for a subset of operations in the transformer model. In one embodiment, the operation to be processed individually is an attention operation of an encoder or a decoder of the transformer model. By selective batching, the inference system can allow batching operations to be performed for a batch of requests with variable input or target length or internal state length to utilize the parallel computation capabilities of hardware accelerators while preventing unnecessary computations that occur for workarounds that restrain the data of a batch of requests to a same length.

A network interface can process a workload request and determine a resource to use to perform the workload request and to generate an executable for execution by the determined resource. A client device or software can determine available resource types. The client device or software can issue a request to perform a workload using a particular resource type. Using telemetry data and performance indicators of available resources, the network interface can select a resource to use to perform the workload. The network interface can translate a workload instruction into a format acceptable by the selected resource and provide the instruction in executable format to the selected resource.

A method includes identifying high-availability jobs and low-availability jobs that demand usage of resources of a distributed system. The method includes determining a first quota of the resources available to low-availability jobs as a quantity of the resources available during normal operations, and determining a second quota of the resources available to high-availability jobs as a quantity of the resources available during normal operations minus a quantity of the resources lost due to a tolerated event. The method includes executing the jobs on the distributed system and constraining a total usage of the resources by both the high-availability jobs and the low-availability jobs to the quantity of the resources available during normal operations.