Embodiments of the present application provide an occupancy information prediction method, an encoder, a decoder, and a storage medium. The occupancy information prediction method comprises: when an encoder encodes geometrical information on the basis of an octree, determining encoding information corresponding to a neighboring node of a node to be predicted, and a distance parameter between a child node of the node to be predicted and the neighboring nodes; wherein the encoding information corresponding to the neighboring node comprises occupancy information; determining an occupancy weight corresponding to the child node of the node to be predicted according to the distance parameter, and the encoding information corresponding to the neighboring node; performing a prediction processing on the child node according to the occupancy weight and a preset occupancy threshold set to obtain a node type corresponding to the child node.

According to one embodiment, an interleaving unit divides a symbol string into first and second symbols. A first coding unit converts the first symbols to first codewords. A first packet generating unit generates first packets including the first codewords. A first request generating unit generates first packet requests including sizes of variable length packets. A second coding unit converts the second symbols to second codewords. A second packet generating unit generates second packets including the second codewords. A second request generating unit generates second packet requests including sizes of variable length packets. A multiplexer outputs a compressed stream including the first and second variable length packets cut out from the first and second packets.

Technologies for dynamically managing resources in disaggregated accelerators include an accelerator. The accelerator includes acceleration circuitry with multiple logic portions, each capable of executing a different workload. Additionally, the accelerator includes communication circuitry to receive a workload to be executed by a logic portion of the accelerator and a dynamic resource allocation logic unit to identify a resource utilization threshold associated with one or more shared resources of the accelerator to be used by a logic portion in the execution of the workload, limit, as a function of the resource utilization threshold, the utilization of the one or more shared resources by the logic portion as the logic portion executes the workload, and subsequently adjust the resource utilization threshold as the workload is executed. Other embodiments are also described and claimed.

Optimized memory usage and management is crucial to the overall performance of a neural network (NN) or deep neural network (DNN) computing environment. Using various characteristics of the input data dimension, an apportionment sequence is calculated for the input data to be processed by the NN or DNN that optimizes the efficient use of the local and external memory components. The apportionment sequence can describe how to parcel the input data (and its associated processing parameters—e.g., processing weights) into one or more portions as well as how such portions of input data (and its associated processing parameters) are passed between the local memory, external memory, and processing unit components of the NN or DNN. Additionally, the apportionment sequence can include instructions to store generated output data in the local and/or external memory components so as to optimize the efficient use of the local and/or external memory components.

Methods of encoding and decoding data are described wherein the encoding method comprises: receiving a data file and dividing the data file or data stream into one or more data blocks, each data block having a predetermined size N and comprising a sequence of data units, e.g. byte values; and, iteratively encoding the data file into a data key based on a first permutation function and a first dictionary of permutation indices, preferably the encoded data file having a total size that is equal to or smaller than the original data file and preferably the data key having a size that is equal to or smaller than size of a data block. Iteratively encoding the data file comprises one or more encoding iterations, wherein each encoding iteration includes: determining a first permutation index defining a permutation to generate the first input data block from a first ordered data block, the generating including providing at least the first input data block to an input of the first permutation function, and the first ordered data block being obtainable by ordering the first input data block; determining a first permutation dictionary index representing a location in the first dictionary in which the first permutation index is stored; generating a first frequency data block defining the number of occurrences for each potential data value in the input data block, preferably determining the number of occurrences for each potential data value in the input data block and ordering the determined occurrences in a sequence of values in a hierarchical order, e.g. increasing or decreasing order of the data value; processing the frequency data block; and determining an encoded data block, the encoded data block comprising the first permutation dictionary index and the processed frequency data block. The encoding method further comprises outputting the data key comprising the one or more encoded data blocks and, optionally, iteration information.

The inventor has conceived, and reduced to practice, a system and method for data compaction using that applies delta encoding methods to entropy encoding methods to improve data compaction of entropy encoding methods under certain conditions and when compacting data having certain characteristics. Delta encoding may be applied to entropy encoding methods to further compact data sets by reducing the number of sourceblocks included in a codebook to those most commonly encountered in data to be encoded and, where mismatches occur during encoding, using delta encoding of bit differences with existing sourceblocks in the codebook rather than adding new sourceblocks to the codebook.

A stream decompression circuit is disclosed. The stream decompression circuit includes a coding length first-in-first-out (FIFO) and a calculation circuit. The coding length FIFO is coupled to a variable length coding (VLC) circuit and used to store a coding length that the VLC circuit codes sub-streams and output a specific number of bits when the coding length accumulates over the specific number of bits. The calculation circuit is coupled between the coding length FIFO and a multiplexer circuit and used to calculate a number of bits required for decompression and output an output multiplex control signal to the multiplexer circuit to control the multiplexer circuit to output the sub-streams according to a specific order.

A method is presented for compressing data of a Rectified Linear Unit (ReLU) function on a graphical processing unit (GPU) employed in a learning process of a deep neural network. The method includes converting an initial data structure including nonzero data and zero data into a compressed data structure including only the nonzero data of the initial data structure as compressed data by generating a nonzero data bitmap region, generating a nonzero data number table region by employing a parallel reduction algorithm, calculating a nonzero data array index per block region of all blocks from the nonzero data number table region by employing a parallel prefix sum scan algorithm, allocating a buffer for the compressed data; and copying the nonzero data from the initial data structure into a nonzero data array region in a compressed data format in parallel.

Techniques are provided for implementing additional compression for existing compressed data. Format information stored within a data block is evaluated to determine whether the data block is compressed or uncompressed. In response to the data block being compressed according to a first compression format, the data block is decompressed using the format information. The data block is compressed with one or more other data blocks to create compressed data having a second compression format different than the first compression format.

An apparatus for encoding directional audio coding parameters comprising diffuseness parameters and direction parameters having a parameter calculator (100) for calculating the diffuseness parameters with a first time or frequency resolution and for calculating the direction parameters with a second time or frequency resolution; and a quantizer and encoder processor (200) for generating a quantized and encoded representation of the diffuseness parameters and the direction parameters.