In some embodiments, a real-time event is detected and context is determined based on the real-time event. An application model is fetched based on the context and meta-data associated with the real-time event, the application model referencing a micro-function and including pre-condition and post-condition descriptors. A graph is constructed based on the micro-function. The micro-function is transformed into micro-capabilities by determining a computing resource for execution of a micro-capability by matching pre-conditions and post-conditions of the micro-capability, and enabling execution and configuration of the micro-capability on the computing resource by providing access in a target environment to an API capable of calling the micro-capability to configure and execute the micro-capability. A request is received from the target environment to execute and configure the micro-capability on the computing resource. The micro-capability is executed and configured on the computing resource, and an output of the micro-capability is provided to the target environment.

In one embodiment, a processor includes: a plurality of cores each comprising a multi-threaded core to concurrently execute a plurality of threads; and a control circuit to concurrently enable at least one of the plurality of cores to operate in a single-threaded mode and at least one other of the plurality of cores to operate in a multi-threaded mode. Other embodiments are described and claimed.

Disclosed are aspects of virtual graphics processing unit (vGPU) scheduling-aware virtual machine migration. Graphics processing units (GPUs) that are compatible with a current virtual GPU (vGPU) profile for a virtual machine are identified. A scheduling policy matching order for a migration of the virtual machine is determined based on a current vGPU scheduling policy for the virtual machine. A destination GPU is selected based on a vGPU scheduling policy of the destination GPU being identified as a best available vGPU scheduling policy according to the scheduling policy matching order. The virtual machine is migrated to the destination GPU.

Systems and methods are disclosed for scheduling threads on a processor that has at least two different core types, such as an asymmetric multiprocessing system. Each core type can run at a plurality of selectable voltage and frequency scaling (DVFS) states. Threads from a plurality of processes can be grouped into thread groups. Execution metrics are accumulated for threads of a thread group and fed into a plurality of tunable controllers for the thread group. A closed loop performance control (CLPC) system determines a control effort for the thread group and maps the control effort to a recommended core type and DVFS state. A closed loop thermal and power management system can limit the control effort determined by the CLPC for a thread group, and limit the power, core type, and DVFS states for the system. Deferred interrupts can be used to increase performance.

Various embodiments, methods and systems for implementing a distributed computing system machine-learning training service are provided. Initially a machine learning model is accessed. A plurality of synthetic data assets are accessed, where a synthetic data asset is associated with asset-variation parameters that are programmable for machine-learning. The machine learning model is retrained using the plurality of synthetic data assets. The machine-learning training service is further configured for executing real-time calls to generate an on-the-fly-generated synthetic data asset such that the on-the-fly-generated synthetic data asset is rendered in real-time to preclude pre-rendering and storing the on-the-fly-generated synthetic data asset. The machine-learning training service further supports hybrid-based machine learning training, where the machine learning model is trained based on a combination of the plurality of synthetic data assets, a plurality of non-synthetic data assets, and synthetic data asset metadata associated with the plurality of synthetic data assets.

Disclosed are various examples for platform independent graphics processing unit (GPU) profiles for more efficient utilization of GPU resources. A virtual machine configuration can be identified to include a platform independent graphics computing requirement. Hosts can be identified as available in a computing environment based on the platform independent graphics computing requirement. The virtual machine can be placed on a host based on a consideration of host priority.

Technologies for providing shared memory for accelerator sleds includes an accelerator sled to receive, with a memory controller, a memory access request from an accelerator device to access a region of memory. The request is to identify the region of memory with a logical address. Additionally, the accelerator sled is to determine from a map of logical addresses and associated physical address, the physical address associated with the region of memory. In addition, the accelerator sled is to route the memory access request to a memory device associated with the determined physical address.

The present disclosure provides a method and system for selecting an optimal edge computing node in an Internet of vehicle (IoV) environment. The method includes: acquiring and analyzing properties of computing tasks of a vehicle in the IoV environment; acquiring and analyzing properties of different edge computing nodes; computing matching degrees between the properties of the computing tasks and the properties of the nodes; analyzing computing demands of different tasks, and assigning weights to different types of matching degrees; and selecting a node having an optimal sum for products of the matching degrees and the weights as an optimal edge computing node to compute each of the computing tasks of the vehicle.

This disclosure relates generally to system and method for molecular property prediction. Typically, message-pooling mechanism employed in molecular property prediction using conventional message passing neural networks (MPNN) causes over smoothing of the node embeddings of the molecular graph. The disclosed system utilizes edge conditioned identity mapping convolution neural network for the message passing phase. In message passing phase, the system computes an incoming aggregated message vector for each node of the plurality of nodes of the molecular graph based on encoded message received from neighboring nodes such that encoded message vector is generated by fusing a node information and an connecting edge information of the set of neighboring nodes of the node. The incoming aggregated message vector is utilized for computing updated hidden state vector of each node. A discriminative graph-level vector representation is computed by pooling the updated hidden state vectors from all the nodes of the molecular graph.

Systems and systems for hybrid centralized distributive scheduling and conflict resolution of multiple scheduler instances that share physical resources in a cloud computing system. The cloud computing system includes a plurality of scheduler instances, a global resource manager (GRM) for high-level resource management and conflict resolution for the scheduler instances, and a plurality of physical hosts. Each physical host has a respective local resource manager (LRM). The scheduler instances are responsible for initially processing of scheduling and resource allocation for resource requests, and proposing candidate physical hosts (and respective resource allocation) for the resource requests to the GRM. The GRM is responsible for conflict resolution through its general conflict resolvers of filtering, sorting and counting. The GRM decides which physical hosts among the candidate physical hosts will run the runtime instances of the resource requests after resolving conflicts among the scheduler instances.