The Fast Fourier Transforms (FFT) are an efficient class of algorithms for digital computation of the Discrete Fourier Transform (DFT). During signal processing, the time domain and frequency domain information of the signal can be collected and analyzed.
FFT Algorithm Accelerator
Overview
Key Features
- Support transform point sizes n = 2^m, m = 4 ~ 12
- Support 16-bit and 32-bit fixed-point data formats
- Support bus max transmission width 64bit
- Support Radix-2 Decimation-In-Time (DIT) and Decimation-In-Frequency (DIF) FFT, IFFT algorithms
- Support Rounding and truncation saturation after the butterfly, to prevent overflow
- Support three types of input data array: imaginary part and real part (IRIR), Real numbers first and imaginary numbers later(RnIn), only real part (_R_R), real part and imaginary part separation(RR,II)
- Support single-port local SRAM cache, and the bus can independently access the SRAM function
- Support acceleration of external functions similar to computational units, such as scale, complex mul, real mul,offset, etc
Technical Specifications
Short description
FFT Algorithm Accelerator
Vendor
Vendor Name