A data compression apparatus includes a lossy compression unit for calculating, in advance, a reference error vector magnitude depending on lossy compression and decompression by removing lower bits from a data signal composed of bitstreams, for setting removal target lower bits satisfying a preset error vector magnitude requirement, and for lossily compressing the data signal by removing the removal target lower bits from the data signal, and a communication unit for transmitting the compressed data signal to a data decompression apparatus.

Various energy efficient data encoding schemes and computing devices are disclosed. In one aspect, a method of transmitting data from a transmitter to a receiver connected by plural wires is provided. The method includes sending from the transmitter on at least one but not all of the wires a first wave form that has first and second signal transitions. The receiver receives the first waveform and measures a first duration between the first and second signal transitions using a locally generated clock signal not received from the transmitter. The first duration is indicative of a first particular data value.

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, determining a target reverberation gain parameter that needs to be encoded in the reverberation gain parameters corresponding to the different subbands of the first channel signal and the second channel signal, generating parameter indication information, where the parameter indication information is used to indicate a subband corresponding to the target reverberation gain parameter, and encoding the target reverberation gain parameter, the parameter indication information, and the downmixed signal to generate a bitstream.

Systems and methods for replicating data are disclosed. Data stored in a compressed form on a source storage array or device can be read and transmitted to a destination storage array or device. The replication of data is achieved without having to decompress the data.

An apparatus comprises a first storage system comprising a plurality of storage devices. The first storage system is configured to participate in a replication process with a second storage system. The first storage system is further configured to identify data to be replicated to the second storage system as part of the replication process, to obtain information characterizing network behavior of at least one network connecting the first storage system to the second storage system, to select a compression method from a set of available compression methods based on the obtained information characterizing the network behavior of said at least one network, to compress the data to be replicated to the second storage system utilizing the selected compression method, and to provide the compressed data to the second storage system.

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.

There is provided a data transmission and compression method arranged to compress and transmit data between an edge device and a remote server. The method comprises collecting data at the edge device, wherein the data is attributed to a plurality of signal signatures; generating a data matrix at the edge device; transforming the data matrix, wherein transforming the data comprises using a wavelet transform; compressing the data, wherein compressing the data comprises utilising an autoencoder; and transmitting the encoded compressed data to the remote server via a communication channel. The method further comprises, at the remote server, decompressing the data utilising an autoencoder; reconstructing the signal signatures using an inverse wavelet transform; and storing the reconstructed data signatures in a datastore on the remote server.

A system on a chip (SoC) includes a first subsystem, a second subsystem and a compression block connected to the first and second subsystems, wherein the compression block includes a decoder and an encoder. The compression block receives spill data generated by a compute element in one of the first and second subsystems, compresses the spill data using the encoder and stores the compressed spill data in a data block in local memory of one of the compute elements.

A method for storage system data aware compression, the method may include pre-compressing data units received by the storage system, by different pre-compression units to provide different pre-compressed versions of the data units; wherein the different pre-compression schemes are associated with different compression schemes, wherein at least some of the different compression schemes are data type specific compression schemes; calculating entropies of the different pre-compressed versions; and selecting a compression scheme out of the different compression schemes based on the entropies of the different pre-compressed versions.

In accordance with some embodiments of the disclosed subject matter, mechanisms (which can, for example, include systems, methods, and media) for low-power encoding of continuous physiological signals are provided. In some embodiments, a system comprises: a physiological sensor; and a remote monitor comprising: a battery; memory storing a k-ary tree including a root with k branches corresponding to k delta values, k nodes at a first depth below the Leads root node each having k branches corresponding to the k delta values the nodes indexed to indicate the lateral position of the node within the depth; a processor programmed to: receive a first sample value from the sensor; receive a second sample value; calculate a difference between the second first sample values; determine that the delta corresponds to a first delta of the k delta values; encode a sequence of deltas based on a depth and node index.