Inside all Zynq All Programmable SoCs, you'll find a dual-core ARM Cortex-A9 MPCore processor with an ARM NEON SIMD architecture extension set in both processors in the Zynq SoC. So why do you need to use the ARM NEON SIMD extension set? That's because you can greatly improve your software performance. You may not see the dramatic acceleration you get from turning mission-critical tasks into Zynq SoC programmable logic, but in many applications you will see some kind of acceleration. The trick is to use the NEON extension set! The secret of this particular decision lies in the NE10 open source library of Project Ne10. The goal of the Ne10 project was to provide a common, efficient set of functions that would greatly optimize the ARM architecture, including the NEON SIMD extension set, and provide unified and test-friendly operations for integration into the application. This open source library supports static linking and dynamic linking, and is modular, so there is no need to add unused library code to the code scripts you write. The following is a list of functions currently available in the NE10 open source library: The following figure shows a set of FFT functions that show the performance comparison of using NEON extensions and Ne10 libraries with vector code generated using NEON inline functions and ARMCC compilers. Ceramic Insulator,Ceramic Electrical Insulators,Ceramic Standoff Insulators,Ceramic Isolators Yixing Guangming Special Ceramics Co.,Ltd , https://www.yxgmtc.com
FIR filter
FIR extractor
FIR insertion complex FFT
Real FFT
Lattice IIR filter
FIR lattice sparse matrix FIR filter dot product vector addition vector subtraction vector accumulation vector division vector set vector length vector normalization vector attribute cross type vector matrix addition matrix subtraction matrix multiplication matrix division matrix vector multiplication matrix transpose matrix determinant Inverse matrix matrix transposed unit matrix 