An arithmetic logic unit, a floating point number processing method, a gpu chip, and an electronic device

Abstract

The present disclosure provides an arithmetic logic unit, including a specific multiplier-adder and a floating-point number control circuit; the specific multiplier-adder is obtained by transforming a single-precision floating-point number multiplier-adder; the floating-point number control circuit is used for receiving Three double-precision floating-point numbers to be multiplied and added, the three double-precision floating-point numbers include two multipliers and one addend, the calculation object of a specific multiplier-adder is determined according to the mantissa of the two multipliers, and the The calculation object is input to the specific multiplier-adder; the specific multiplier-adder is used to multiply the input calculation object, and return the calculation result to the floating-point number control circuit; the floating-point number control circuit is used to receive the specific multiplication and addition The calculation result of the device is obtained, and the product result of the two multipliers is obtained according to the calculation result; the product result is added with the addend to obtain the multiplication and addition result of the three double-precision floating-point numbers, so The result of the multiplication and addition is a double-precision floating-point number.

Description

technical field [0001] The present disclosure relates to the field of computer technology, and in particular, to an arithmetic logic unit, a floating-point number processing method, a GPU chip, and an electronic device. Background technique [0002] Graphics processor GPU chip is a microprocessor used to process images and graphics-related operations. It needs to have strong floating-point calculation capabilities. Among them, the most common and commonly used calculation for floating-point calculations is multiplication. Addition calculation, with the continuous development of technology, the graphics processor GPU is currently required to have the computing ability to multiply and add double-precision floating-point numbers. However, the current graphics processors in the industry only have the ability to multiply and add single-precision floating-point numbers. If If the GPU is required to have the computing capability of multiplying and adding double-precision floating-p...

AI Technical Summary

Problems solved by technology

[0002] Graphics processor GPU chip is a microprocessor used to process images and graphics-related operations. It needs to have a strong ability to calculate floating-point numbers. Among them, the most common and commonly used calculation for floating-point numbers is multiplication. Addition calculation, with the continuous development of technology, graphics processor GPU is now required to have the ability to multiply and add double-precision floating-point numbers. However, the current graphics processors in the industry only have the ability to multiply and add single-precision floating-point numbers. If If the GPU is required to have the ability to multiply and add double-precision floating-point numbers, it is necessary to design and develop a multiplier-accumulator for double-precision floating-point numbers, and deploy the developed multiplier-accumulator for double-precision floating-point numbers on the GPU chip. This method will face higher design difficulty, and also requires a longer development cycle. In addition, deploying the complete double-precision floating-point multiply-adder on the GPU chip also greatly occupies the area of ​​​​the GPU chip.

Images