A computer-implemented method of selecting a power-optimal compression scheme for transmitting digital control signals from a classical interface of a quantum computer to a quantum processing unit (QPU) of the quantum computer is disclosed. The method involves receiving static and dynamic power consumption values associated with operations performable by the QPU; determining compression schemes implementable by the QPU; calculating total power consumption values associated with receiving and decompressing a representative control signal at the QPU using the compression schemes; and selecting the compression scheme having the lowest total power consumption value. A corresponding method for transmitting control signals from a classical interface of the quantum computer to the QPU is also disclosed in which a compressed control signal is transmitted from the classical interface to the QPU with one or more delays.

A multi-channel signal encoding method includes determining a downmixed signal of a first channel signal and a second channel signal in a multi-channel signal, and reverberation gain parameters corresponding to different subbands of the first channel signal and the second channel signal, where the obtained reverberation gain parameters belong to at least two reverberation gain parameter groups. The method further includes selecting, from the at least two reverberation gain parameter groups, a target reverberation gain parameter group. The method further includes generating parameter indication information, where the parameter indication information indicates the target reverberation gain parameter group. The method further includes encoding reverberation gain parameters corresponding to the target reverberation gain parameter group, the parameter indication information, and the downmixed signal to obtain a bitstream.

Compressed domain processors configured to perform operations on data compressed in a format that preserves order. The Compressed domain processors may include operations such as addition, subtraction, multiplication, division, sorting, and searching. In some cases, compression engines for compressing the data into the desired formats are provided.

A data compression method comprises encoding groups of data items by generating, for each group, header data comprising h-bits and a plurality of body portions each comprising b-bits and each body portion corresponding to a data item in the group. The value of h may be fixed for all groups and the value of b is fixed within a group, wherein the header data for a group comprises an indication of b for the body portions of that group. In various examples, b=0 and so there are no body portions. In examples where b is not equal to zero, a body data field is generated for each group by interleaving bits from the body portions corresponding to data items in the group. The resultant encoded data block, comprising the header data and, where present, the body data field can be written to memory.

Methods, apparatus, systems and articles of manufacture are disclosed for sparse tensor storage for neural network accelerators. An example apparatus includes sparsity map generating circuitry to generate a sparsity map corresponding to a tensor, the sparsity map to indicate whether a data point of the tensor is zero, static storage controlling circuitry to divide the tensor into one or more storage elements, and a compressor to perform a first compression of the one or more storage elements to generate one or more compressed storage elements, the first compression to remove zero points of the one or more storage elements based on the sparsity map and perform a second compression of the one or more compressed storage elements, the second compression to store the one or more compressed storage elements contiguously in memory.

In an approach for compressing a neural network, a processor receives a neural network, wherein the neural network has been trained on a set of training data. A processor receives a compression ratio. A processor compresses the neural network based on the compression ratio using an optimization model to solve for sparse weights. A processor re-trains the compressed neural network with the sparse weights. A processor outputs the re-trained neural network.

The present disclosure discloses example operation accelerators and compression methods. One example operation accelerator performs operations, including storing, in a first buffer, first input data. In a second buffer, weight data can be stored. A computation result is obtained by performing matrix multiplication on the first input data and the weight data by an operation circuit connected to the input buffer and the weight buffer. The computation result is compressed by a compression module to obtain compressed data. The compressed data can be stored into a memory outside the operation accelerator by a direct memory access controller (DMAC) connected to the compression module.

A system and method for distributed node-based data compaction. The system uses machine learning on data chunks to generate codebooks which compact the data to be stored, processed, or sent with a smaller data profile than uncompacted data. The system uses a data compaction in an existing blockchain fork or implemented in a new blockchain protocol from which nodes that wish to or need to use the blockchain can do so with a reduced storage requirement. The system uses network data compaction across all nodes to increase the speed of and decrease the size of a blockchain's data packets. The system uses data compaction firmware to increase the efficiency at which mining rigs can computationally validate new blocks on the blockchain. The system can be implemented using any combination of the three data compaction services to meet the needs of the desired blockchain technology.

An information processing apparatus includes: a processor; and a processing circuit coupled to the processor, wherein the processing circuit is configured to: generate compressed data by compressing send data; and determine whether to transmit the compressed data or the send data before the compression to a network, based on a size of the compressed data, and wherein the processor is configured to transmit the compressed data or the send data before the compression to the network, based on a result of the determination.

Embodiments described herein provide systems and methods for sharing encoder output of video streams. In a particular embodiment, a method provides determining video profiles for each of a plurality of devices. The method further provides determining if two or more of the video profiles are similar by determining if parameters associated with each video profile differ by less than a threshold value. The method further provides transmitting a video stream encoded in a single format to the devices if they have similar profiles and transmitting video streams encoded in different formats to the devices if they do not have similar profiles.