Abstract: Apparatus and method for processing attention operations. For example, a plurality of queues are allocated in memory, the plurality of queues corresponding to a plurality of heterogeneous processor types; a first set of nodes of an attention graph are submitted which have data dependencies resolved in accordance with a dependency tracking data structure, to the plurality of queues; the first set of nodes are dispatched for execution from the selected queues, starting with a node corresponding to a processor of one of the heterogeneous processor types which is associated with a relatively longer execution time than other nodes executed on other heterogeneous processor types; updating the dependency tracking data structure to reflect corresponding newly-resolved data dependencies; and submitting a second set of nodes of the attention graph associated with the corresponding newly-resolved dependencies in accordance with the dependency tracking data structure, to the plurality of queues.

Abstract: A system is disclosed that has at least one input to receive at least one model and data to be processed with the at least one model, and at least one circuit configured to perform: in response to receiving, via the input, a first amount of data to be processed with the at least one model, based on configuration data indicating that when input data of the first amount is to be processed with the at least one model the input data should be supplemented with additional data to yield a second amount of data to be processed, generating additional data to supplement the first amount of data and yield aggregated data that has the second amount of data, the second amount of data being larger than the first amount of data, processing the aggregated data with the at least one model. Various other methods and systems are also disclosed.

Abstract: Methods and systems are disclosed for distribution of a workload among nodes of a NUMA architecture. Techniques disclosed include receiving the workload and data batches, the data batches to be processed by the workload. Techniques disclosed further include assigning workload processes to the nodes according to a determined distribution, and, then, executing the workload according to the determined distribution. The determined distribution is selected out of a set of distributions, so that the execution time of the workload, when executed according to the determined distribution, is minimal.

Abstract: An example apparatus includes interface circuitry, machine-readable instructions, and at least one processor circuit to be programmed by the machine-readable instructions to distribute a computing task across one or more cores based on a batch dimension, execute a row pooling operation to generate a first intermediate data output based on the batch dimension and a first pooling size, execute a column pooling operation to produce a second intermediate data output based on the batch dimension and a second pooling size, and generate a final output based on the first intermediate data output and the second intermediate data output.

Abstract: Methods and systems are disclosed for distribution of a workload among nodes of a NUMA architecture. Techniques disclosed include receiving the workload and data batches, the data batches to be processed by the workload. Techniques disclosed further include assigning workload processes to the nodes according to a determined distribution, and, then, executing the workload according to the determined distribution. The determined distribution is selected out of a set of distributions, so that the execution time of the workload, when executed according to the determined distribution, is minimal.

Abstract: Devices, methods, and systems for determining N-dimensional MaxPool or AvgPool for a M-dimensional input array. For each of N dimensions, in order from highest to lowest dimension i: the M dimensional input array is decomposed into 1 dimensional (1D) input arrays in the ith dimension, 1D MaxPool or AvgPool is performed on each of the 1D input arrays in the ith dimension to generate 1D output arrays in the ith dimension, and the M dimensional input array is recomposed from the 1D output arrays in the ith dimension to update the M-dimensional input array. In MaxPool, the updated M-dimensional input array is output as an M-dimensional output array. In AvgPool, each element of the updated M-dimensional input array is divided by a kernel size to form the M-dimensional output array.

Abstract: A broadband communication system with improved latency is disclosed. The system employs acceleration of secure web-based communications over a satellite communication network. In accordance with aspects of the invention, secure protocol acceleration is employed such that required protocol signals transmitted from a computer employing a web browser may be intercepted by a remote terminal. To insure that the browser will continue transmitting data, the remote terminal generates required acknowledgment and security signals to continue the secure communication, which may then transmitted back to the computer. Meanwhile, the received protocol signals may be converted by the remote terminal for transmission through the satellite communications system in a format appropriate for that communication medium. Aspects of the invention further include a hub or similar device for communicating with the satellite communications system.

Abstract: A broadband communication system with improved latency is disclosed. The system employs acceleration of secure web-based communications over a satellite communication network. In accordance with aspects of the invention, secure protocol acceleration is employed such that required protocol signals transmitted from a computer employing a web browser may be intercepted by a remote terminal. To insure that the browser will continue transmitting data, the remote terminal generates required acknowledgment and security signals to continue the secure communication, which may then transmitted back to the computer. Meanwhile, the received protocol signals may be converted by the remote terminal for transmission through the satellite communications system in a format appropriate for that communication medium.

Abstract: A network appliance may be configured to transmit data over a multiple networks from one or more client applications, for example, enterprise business applications. The network appliance may receive and analyze information regarding the multiple different networks transmitting over different transmission media, e.g., terrestrial and wireless, and to switch a client application from a primary network to a backup network based on a detected change in network availability. For example, the determination to switch transmission networks may be based on an outage of a primary network, or based on an efficiency or cost analysis, along with other factors. The network appliance may be in communication with an appliance network server, for example, to receive and compile statistics and remotely reconfigure a plurality of network appliances installed in remote locations.