Abstract: A method for sending data across processors to combine the data on the processors is described. In one embodiment, a method includes receiving a set of data at a set of processors configured in an asymmetric or symmetric tree topology including a root and one or more leaves. Target portions of the set of data are assigned to processors of the set of processors based on a number of child processors that are connected to a parent node. The method includes sending iteratively apportioned combined data between child processors sharing the same parent node in each branch of the tree topology starting from the one or more leaves and increasing levels in the tree topology until reaching the root. The method also includes sending the combined data between child processors from one branch to child processors in at least one other branch.

Abstract: A pretrained model is selected to operate in an augmented model configuration with a submodel. The submodel is trained using training data corresponding to a second domain, whereas the pretrained model is trained to operate on data of a first domain. The pretrained model is augmented, to form the augmented model configuration, with the submodel, by combining a first feature map being output from a layer in the pretrained model with a second feature map being output from a layer in the submodel. The combining forms a combined feature map. The combined feature map is input into a different layer in the submodel.

Abstract: A method for executing multi-directional reduction algorithms includes identifying a set of nodes, wherein a node includes at least one data element, creating a set of partitions including one or more data elements from at least two nodes, wherein the at least two nodes are arranged in a single direction with respect to the positioning of the set of nodes, executing a reduction algorithm on the data elements within the created set of partitions, creating an additional set of partitions including one or more data elements from at least two nodes, wherein the at least two nodes are arranged in a different direction with respect to the positioning of the set of nodes, executing a reduction algorithm on the data elements within the created additional set of partitions, and providing a set of reduced results corresponding to the at least one data element.

Abstract: A computer-implemented method comprises linking a private AI model to a public AI model to thereby form a combined AI model comprising the private AI model and the public AI model; and training the combined AI model with private samples while keeping the public AI model fixed so that only the private AI model is trained with the private samples.

Abstract: A computer-implemented method comprises: inputting into an autoencoder sets of input samples, each of the sets of input samples comprising: a reference input sample of a reference dataset and one or more target input samples of one or more target datasets, the autoencoder being trained using the reference dataset. The autoencoder generates a respective set of outputs for each set of the input samples to thereby form one or more respective sets of outputs, each of the one or more sets of outputs comprising the reference output and the one or more target outputs for a respective set of input samples; and determining the similarity of each of the one or more target datasets to the reference dataset by comparing each of the one or more target outputs to respective target input samples of each of the sets of input samples.

Abstract: A computer-implemented method generates a hamming code based target label for each class of a dataset in which hamming distance between the target labels in the dataset is maximized and trains a convolutional neural network with the hamming codes based target label to thereby produce a trained AI model. The confusability between classes of the dataset is determined using a confusion matrix. The hamming distances of classes of the dataset that are determined to be more confusable are set to higher values than the hamming distances of classes of the dataset that are determined to be less confusable.

Abstract: Apparatuses and Methods for sorting a data set. A data storage is divided into a plurality of buckets that is each associated with a respective key value. A plurality of stripes is identified in each bucket. A plurality of data stripe sets is defined that has one stripe within each respective bucket. A first and a second in-place partial bucket radix sort are performed on data items contained within the first and second data stripe sets, respectively, using an initial radix. Incorrectly sorted data items in the first bucket are grouped by a first processor and incorrectly sorted data items in the second bucket are grouped by a second processor into a respective incorrect data item group within each bucket. A radix sort is then performed using the initial radix on the items within the respective incorrect data item group. A first level sorted output is produced.

Abstract: A method for sending data across processors to combine the data on the processors is described. In one embodiment, a method includes receiving a set of data at a set of processors configured in an asymmetric or symmetric tree topology including a root and one or more leaves. Target portions of the set of data are assigned to processors of the set of processors based on a number of child processors that are connected to a parent node. The method includes sending iteratively apportioned combined data between child processors sharing the same parent node in each branch of the tree topology starting from the one or more leaves and increasing levels in the tree topology until reaching the root. The method also includes sending the combined data between child processors from one branch to child processors in at least one other branch.

Abstract: Matrix factorization based gradient compression may be applied to an allreduce operation to improve efficiency including the elimination of unnecessary meta data while maintaining accuracy in training of deep learning (DL) of Artificial Intelligence. This compression may include generating a predetermined matrix and a degree of data compression k as a dimension of the predetermined matrix for a plurality of computing nodes. Each computing node may receive a corresponding matrix of matrices to be allreduced, and each corresponding matrix may be decomposed into a plurality of non-fixed matrices and the predetermined matrix. The plurality of non-fixed matrices may be summed to provide an optimized matrix, which may be multiplied by the predetermined matrix to provide a result matrix. The optimized matrix may be designated as a predetermined matrix. These operations may be repeated until all of the matrices received by the computing nodes have been allreduced.

Abstract: Various embodiments are provided for compression for allreduce in a deep learning by one or more processors in a computing system. A gradient may be compressed for synchronization in a data parallel deep neural network training for allreduce by sharing a consensus vector between each node in a plurality of nodes to ensure identical indexing in each of the plurality of nodes prior to performing sparse encoding.

Abstract: An iterative graph algorithm accelerating method, system, and computer program product, include recording an order of access nodes in a memory layout, reordering the access nodes in the memory layout in accordance with the recorded order, and updating edge information of the reordered access nodes.

Abstract: Various embodiments are provided for using pessimistic scheduling for topology optimized workload placement by a processor in a computing environment. An excessive amount of computing resources may be requested (e.g., a pessimistic request) to execute a workload as compared to a required amount of the computing resources to execute the workload. The workload may be assigned to a selected configuration of the excessive amount of computing resources and releasing a remaining amount of the excessive amount of computing resources.

Abstract: Embodiments for pipelining multi-directional reduction by one or more processors in a computing system. One or more reduce scatter operations and one or more all-gather operations may be assigned to each of a plurality of independent networks. The one or more reduce scatter operations and the one or more all-gather operations may be sequentially executed in each of the plurality of independent networks according to a serialized execution order and a defined time period.

Abstract: An information processing system, computer readable storage medium, and method for accelerated radix sort processing of data elements in an array in memory. The information processing system stores an array of data elements in a buffer memory in an application specific integrated circuit radix sort accelerator. The array has a head end and a tail end. The system radix sort processing, with a head processor, data elements starting at the head end of the array and progressively advancing radix sort processing data elements toward the tail end of the array. The system radix sort processing, with a tail processor, data elements starting at the tail end of the array and progressively advancing radix sort processing data elements toward the head end of the array, the tail processor radix sort processing data elements in the array contemporaneously with the head processor radix sort processing data elements in the array.

Abstract: Various embodiments are provided for decentralized distributed deep learning by one or more processors in a computing system. Asynchronous distributed training of one or more machine learning models may be performed by generating a list of neighbor nodes for each node in a plurality of nodes and creating a first thread for continuous communication according to a weight management operation and a second thread for continuous computation of a gradient for each node. One or more variables are shared between the first thread and the second thread.

Abstract: A first quicksort is performed in parallel across pairs of partitions of a dataset assigned to respective ones of available processors, including swapping elements of a first partition of a given one of the pairs that are larger than a pivot with elements of a second partition of the given pair that are smaller than the pivot. A second quicksort is performed in parallel across those partitions having elements left unsorted by the first quicksort, and first misplaced elements from a first side of the dataset corresponding to the first partition are swapped with second misplaced elements from a second side of the dataset corresponding to the second partition to produce a first dataset having elements equal to or lower than the pivot and a second dataset having elements equal to or higher than the pivot.

Abstract: Techniques that facilitate model support in deep learning are provided. In one example, a system includes a graphics processing unit and a central processing unit memory. The graphics processing unit processes data to train a deep neural network. The central processing unit memory stores a portion of the data to train the deep neural network. The graphics processing unit provides, during a forward pass process of the deep neural network that traverses through a set of layers for the deep neural network from a first layer of the set of layers to a last layer of the set of layers that provides a set of outputs for the deep neural network, input data for a layer from the set of layers for the deep neural network to the central processing unit memory.

Abstract: A batched Cholesky decomposition method, system, and non-transitory computer readable medium for a Graphics Processing Unit (GPU), include mirroring a second problem matrix of a second problem to a first problem matrix of a first problem as paired matrices and shifting the second problem matrix by N+1 and combining the first problem matrix and the mirrored second problem matrix into one matrix of (N+1)×N by merging the first problem matrix and the mirrored second problem matrix. The first problem matrix and the second problem matrix are symmetric and positive definite matrices.

Abstract: A batched Cholesky decomposition method, system, and non-transitory computer readable medium for a Graphics Processing Unit (GPU), include mirroring a second problem matrix of a second problem to a first problem matrix of a first problem as paired matrices and shifting the second problem matrix by N+1 and combining the first problem matrix and the mirrored second problem matrix into one matrix of (N+1)×N, where the first problem shared memory comprises regular intervals, where the second problem shared memory is continuous, and where the GPU performs batched dense Cholesky decomposition with the one matrix from the combining to accelerate the Cholesky decomposition.

Abstract: Provided are an X-ray detector, an X-ray system including the X-ray detector, and a method of operating the X-ray system. The X-ray detector includes a shock-detecting sensor configured to detect a shock applied to the X-ray detector. By applying a shock-absorbing member to the shock-detecting sensor, it is possible to attenuate the shock applied to the X-ray detector and expand a range of shock values measurable by the shock-detecting sensor. The X-ray system provides information about the shock applied to the X-ray detector to an external server that stores and accumulates the information for later use.