A multipart encoded representation or data type may be particularly useful for limited-capacity processors, such as those used in at-memory or single-instruction, multiple data (SIMD) devices. To encode a source number, the source number is quantized as a first binary representation defined by a first set of exponent bits and a first set of mantissa bits to obtain a first encoded part. A difference between the source number and a nearest dequantization of the first encoded part is computed. The difference is quantized as a second binary representation defined by a second set of exponent bits and a second set of mantissa bits to obtain a second encoded part. The first encoded part and the second encoded part are stored as an encoded representation of the source number. A computational operation may be performed using the encoded representation. The encoded representation may be decoded using an inverse process.

This application provides a data processing method and apparatus, and relates to the field of data compression. The method includes: obtaining to-be-compressed data; sequentially performing n times of preset processing on the to-be-compressed data to obtain preprocessed data, where the preset processing includes: performing a differential operation on rows of a to-be-operated matrix, or performing a differential operation on columns of the to-be-operated matrix, where the to-be-operated matrix is a matrix obtained through a previous time of the preset processing, or the to-be-operated matrix is a matrix formed by the to-be-compressed data; and compressing the preprocessed data through entropy encoding, to obtain compressed data.

A multipart encoded representation or data type may be particularly useful for limited-capacity processors, such as those used in at-memory or single-instruction, multiple data (SIMD) devices. To encode a source number, the source number is quantized as a first binary representation defined by a first set of exponent bits and a first set of mantissa bits to obtain a first encoded part. A difference between the source number and a nearest dequantization of the first encoded part is computed. The difference is quantized as a second binary representation defined by a second set of exponent bits and a second set of mantissa bits to obtain a second encoded part. The first encoded part and the second encoded part are stored as an encoded representation of the source number. A computational operation may be performed using the encoded representation. The encoded representation may be decoded using an inverse process.