Floating-point accumulation/gradual decrease operational method with floating-point precision maintaining function

Abstract

The invention discloses a floating-point accumulation / gradual decrease operational method with a floating-point precision maintaining function. The implementation steps are as follows: (1) increasing the widths of a match exponent register, an expansion result register and a binary adder and decomposing the operation into a floating-point addition / subraction operation and a floating-point accumulation / gradual decrease command; (2) executing the floating-point addition / subraction operation, saving the operation result in the expansion result register, normalizing and rounding the operation result, and then writing the operation result into a target floating-point register; and (3) executing the floating-point accumulation / gradual decrease command, during each time of execution of the floating-point accumulation / gradual decrease command, reading the first operand from the floating-point register and the second operand from the expansion result register, then returning to execute the step (2), and after execution of the last floating-point accumulation / gradual decrease command, outputting data in the target floating-point register. The floating-point accumulation / gradual decrease operational method has the advantages of high floating-point precision, easiness in implementing hardware logic, wide range of application and flexibility and convenience in use.

Description

technical field [0001] The invention relates to a functional unit supporting floating-point addition / subtraction operations in a microprocessor system structure, in particular to a floating-point accumulation / subtraction operation method with the function of maintaining floating-point precision. Background technique [0002] Floating point numbers are a way of representing data in a fixed format. Floating point numbers are represented as figure 1 As shown, it specifically includes the sign bit, the floating-point number exponent field, and the floating-point number mantissa field. The sign bit is used to indicate that the data is greater than zero or less than zero; the floating-point number exponent field is used to determine the position of the decimal point of the data; the floating-point number mantissa field is used to represent the valid digits of the data. However, due to the constraints of computer word length and other conditions, floating-point numbers usually ca...

AI Technical Summary

Problems solved by technology

However, in the floating-point addition and subtraction of the prior art, the data exceeding the bit width of the floating-point number will be truncated in the process of order alignment, and a truncation error will be generated in the process

The normalization and rounding processing after the operation will be truncated according to the bit width of the final result, and the mantissa will be processed according to the rounding method, thus truncation or carry will further generate truncation and rounding errors

Images