A method for computational storage may include storing, at a storage device, two or more portions of data, wherein a first one of the two or more portions of data comprises a first fragment of a record and a second one of the two or more portions of data comprises a second fragment of the record, and performing, by the storage device, an operation on the first and second fragments of the record. The method may further include performing, by the storage node, a second operation on first and second fragments of a second record. The operation may include a data selection operation, and the method may further include sending a result of the data selection operation to a server. The method may further include sending a result of a first data selection operation to a server.

A method for computational storage may include storing, at a storage device, a first portion of data, wherein the first portion of data may include a first fragment of a record, and a second portion of data may include a second fragment of the record, and appending the second fragment of the record to the first portion of data. The method may further include performing, at the storage device, an operation on the first and second fragments of the record. The method may further include determining that the first portion of data may include a first fragment of a record, and a second portion of data may include a second fragment of the record, wherein appending the second fragment of the record to the first portion of data may include appending, based on the determining, the second fragment of the record to the first portion of data.

A method for computational storage may include receiving, at a storage device, a modified version of a portion of data, generating, at the storage device, a restored portion of data from the modified version of the portion of data, and performing, at the storage device, an operation on the restored portion of data. The method may further include receiving, at the storage device, a request to perform the operation on the portion of data. The generating may include decompressing the modified version of the portion of data. The generating may include decrypting the modified version of the portion of data. The method may further include sending, from the storage device, a result of the operation on the restored portion of data. The operation may include a filtering operation. The operation may include a scanning operation. The method may further include dividing data to generate the portion of data.

Technologies for providing shared memory for accelerator sleds includes an accelerator sled to receive, with a memory controller, a memory access request from an accelerator device to access a region of memory. The request is to identify the region of memory with a logical address. Additionally, the accelerator sled is to determine from a map of logical addresses and associated physical address, the physical address associated with the region of memory. In addition, the accelerator sled is to route the memory access request to a memory device associated with the determined physical address.

An electronic device that includes a decompression engine that includes N decoders and a decompressor decompresses compressed input data that includes N streams of data. Upon receiving a command to decompress compressed input data, the decompression engine causes each of the N decoders to decode a respective one of the N streams from the compressed input data separately and substantially in parallel with others of the N decoders. Each decoder outputs a stream of decoded data of a respective type for generating commands associated with a compression standard for decompressing the compressed input data. The decompressor next generates, from the streams of decoded data output by the N decoders, commands for decompressing the data using the compression standard to recreate the original data. The decompressor next executes the commands to recreate the original data and stores the original data in a memory or provides the original data to another entity.

A system comprises an encoder configured to entropy encode a bitstream comprising both compressible and non-compressible symbols. The encoder parses the bitstream into a compressible symbol sub-stream and a non-compressible sub-stream. The non-compressible symbol sub-stream bypass an entropy encoding component of the encoder while the compressible symbol sub-stream is entropy encoded. When a quantity of bytes of entropy encoded symbols and bypass symbols is accumulated a chunk of fixed or known size is formed using the accumulated entropy encoded symbol bytes and the bypass bytes without waiting on the full bitstream to be processed by the encoder. In a complementary manner, a decoder reconstructs the bitstream from the packets or chunks.

A system and method to compress application control data, such as weights for a layer of a convolutional neural network, is disclosed. A multi-core system for executing at least one layer of the convolutional neural network includes a storage device storing a compressed weight matrix of a set of weights of the at least one layer of the convolutional network and a decompression matrix. The compressed weight matrix is formed by matrix factorization and quantization of a floating point value of each weight to a floating point format. A decompression module is operable to obtain an approximation of the weight values by decompressing the compressed weight matrix through the decompression matrix. A plurality of cores executes the at least one layer of the convolutional neural network with the approximation of weight values to produce an inference output.

An integrated circuit includes a compression accelerator to process a request from software to compress source data into an output file. The compression accelerator includes early-abort circuitry to provide for early abort of compression operations. In particular, the compression accelerator uses a predetermined sample size to compute an estimated size for a portion of the output file. The sample size specifies how much of the source data is to be analyzed before computing the estimated size. The compression accelerator also determines whether the estimated size reflects an acceptable amount of compression, based on a predetermined early-abort threshold. The compression accelerator aborts the request if the estimated size does not reflect the acceptable amount of compression. The compression accelerator may complete the request if the estimated size reflects the acceptable amount of compression. Other embodiments are described and claimed.

A hardware decompression acceleration engine including: an input buffer for receiving to-be-decompressed data from a software layer of a host computer; a decompression processing unit coupled to the input buffer for decompressing the to-be-decompressed data, the decompression processing unit further receiving first and second flags from the software layer of the host computer, wherein the first flag is indicative of a location of the to-be-decompressed data in a to-be-decompressed data block and the second flag is indicative of a presence of an intermediate state; and an output buffer for storing decompressed data from the decompression processing unit.

An accelerator device may access an input data chunk to be compressed by the accelerator device. The accelerator device may access an entropy value for the input data chunk. The accelerator device may compress the input data chunk or return an indication that the input data chunk will not be compressed based on the entropy value and an entropy threshold.