The implementation of high-precision floating transcendental functions on FPGA based on the CORDIC algorithm
Xiao Li
Abstract
Xiao Li
Abstract
This paper presents a novel architecture of input and output floating-point execution unit,which can converts data from the format of CORDIC algorithm system to the format accepted by processors.Two different units of angle can be directly supported by the floating-point CORDIC module,the wide angle even above 360° also can be directly supported by the hardware module. Finally,we accomplish the hardware module by adding custom instruction on the Nios Ⅱ processor system of Altera corpora- tion.The floating-point CORDIC module's correctness is proved by C program running in the Nios II processor.
OpenAlex reports 2 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
This paper presents a novel architecture of input and output floating-point execution unit,which can converts data from the format of CORDIC algorithm system to the format accepted by processors.Two different units of angle can be directly supported by the floating-point CORDIC module,the wide angle even above 360° also can be directly supported by the hardware module. Finally,we accomplish the hardware module by adding custom instruction on the Nios Ⅱ processor system of Altera corpora- tion.The floating-point CORDIC module's correctness is proved by C program running in the Nios II processor.
Key concepts: CORDIC, Computer science, Field-programmable gate array, Correctness, Computer hardware, Nios II, Floating point, Point (geometry)