Fixed-point data screening circuit

Abstract

The invention discloses a fixed-point data screening circuit. The fixed-point data screening circuit comprises a rising edge detection circuit, an input circuit, a subtracter and an output circuit, wherein to-be-screened data A, address data B of the to-be-screened data in a memory and a threshold value DR0 during execution of a threshold value comparison instruction are used as input data; the data A is input into the minuend input end of the subtracter, the DR0 or the last piece of data A is input into the subtrahend end of the subtracter, and the output end of the subtracter is connected with the rising edge detection circuit; and the output end of the rising edge detection circuit and the output end of the input circuit are both connected with the output circuit. Through the fixed-point data screening circuit, a large amount of fixed-point data can be screened in a data stream mode, and data or corresponding addresses meeting designated conditions can be given. When the screening circuit is designed in an ALU, screening of fixed-point data can be quickly completed, design of a software program is simplified, execution energy consumption of a processor is lowered, and the defect that an ordinary data comparison instruction can only be used for comparing a small amount of data (generally two pieces of data) is overcome.

Description

【Technical field】 [0001] The invention belongs to the technical field of digital circuits and relates to a fixed-point data screening circuit. 【Background technique】 [0002] In a digital signal processing chip, it is often necessary to screen out the data whose number meets a certain size relationship from a large number of data, and give the data value and the address corresponding to the data that meet the conditions. Chips currently on the market seldom directly provide hardware support for this data filtering function. If it is necessary to find data satisfying a certain size relationship from a large amount of data, it is necessary to use multiple instructions to form an instruction set, and use several cycles to complete the data comparison. In the implementation process of this method, the resource overhead is relatively large, and the instruction execution A long time. Common processors include data comparison instructions. If the comparison of fixed-point data is...

AI Technical Summary

Problems solved by technology

In this way, in order to complete the screening of a set of data, it needs to occupy more register resources, and the code set formed by multiple instructions needs to occupy a large program memory space, and it needs to continuously execute value fetching, decoding, and execution. Waiting for operations, the code execution time is longer, and it will cause more energy consumption

Images