Patent
Fixed point and floating point converter, processor, method and storage medium
Reads0
Chats0
TLDR
In this paper, a fixed point and floating point converter can be embedded in a cloud server to improve the conversion speed of a fixed-point and a floating-point in the cloud server.Abstract:
The embodiment of the invention discloses a fixed point and floating point converter, a processor, a method and a storage medium. The fixed point and floating point converter can be embedded in a cloud server to improve the conversion speed of a fixed point and a floating point in the cloud server. An input module inputs target fixed point data into a left search module and a right search module;the left search module starts from the decimal point position of the target fixed point data and performs integer preamble value search along the integer direction; the rightward search module startsfrom the decimal point position and carries out decimal leading value search along the decimal direction; if the integer preamble value is found by the left search module, a floating point module acquires a first exponential deviation value, and generates floating point data according to the sign bit of the target fixed point data and the first exponential deviation value; and if the integer preamble value is not found, the floating point module acquires a second exponential offset value, and generates floating point data according to the sign bit and the second exponential offset value. By adopting the converter and the method, the universality of conversion from the fixed-point data into the floating-point data can be improved.read more
References
More filters
Patent
System and method for data management
TL;DR: In this article, a data management system that efficiently uses data areas and significantly increases search efficiency is presented, where coordinate data 1 is stored in an index 2 and an index generating algorithm is used to generate the index from the coordinate data.
Patent
Apparatus and method for performing arithmetic operations to accumulate floating-point numbers
TL;DR: In this paper, an apparatus and method for performing arithmetic operations to accumulate floating-point numbers is presented, which comprises execution circuitry to perform arithmetic operations, and decoder circuitry to decode a sequence of instructions.
Patent
High-speed data format conversion part with SIMD (single instruction multiple data) and conversion method thereof
TL;DR: In this paper, a high-speed data format conversion part with SIMD (single instruction multiple data) and a conversion method thereof is presented, which belongs to the field of computer technologies.
Patent
Chip comprising multiply-accumulate module, terminal and control method
TL;DR: In this paper, the authors proposed a chip comprising a multiply-accumulate module, a terminal and a control method, and relates to the field of chips, which includes a multiplication and accumulation module, and the multiplication module comprises a fixed point general unit, a floating point special unit and an output selection unit.
Patent
Device and method for generating floating point constant
TL;DR: In this paper, a device for generating a floating point constant consisting of a decoding unit and a floating-point constant generating unit is presented, where the decoding unit is used for decoding a floatingpoint instruction to generate an instruction indication symbol and an m-bit binary immediate value.