high performance computing on graphics processing units: hgpu.org

With the development of satellite remote sensing technology, satellite remote sensing data obtained by the amount will increase rapidly. Consequently, the process of Wallis transformation is faced with such challenges as large data size, high intensity, high computational complexity and large computational quantity, and so on. A fast algorithm and efficient implementation of Wallis filtering based on Compute Unified Device Architecture (CUDA) is proposed. The parallel hardware architecture and software development process of multiple graphic processing unit (multi-GPU) is introduced firstly. On the basis of the parallel architecture and hardware characteristic of GPU, some algorithms are focused on computation and optimization of multiplicative coefficients, additive coefficients and updated gray values of the image pixels to improve the computing speed, and the shared memory is used to improve the data access efficiency. The experimental results clearly show that, in the same image quality, the average acceleration rate is 107 times, and the algorithm on multi-GPU can achieve better performance.