Mieda, Etsuko (NRC-Herzberg), Fung, Jeffrey (University of California, Berkeley), Veran, Jean-Pierre (NRC-Herzberg)
We modify the pyramid wavefront sensor simulation code PASSATA to enable simulations for wavefront sensing with extended objects and in broadband. Our modifications make use of Graphics Processing Unit (GPU) with parallel algorithms. We compare the computational time among a variety of CPUs and GPUs, and find that our new GPU implementation improves the Fast Fourier Transfer speed by a factor of 150, and as a result, the entire wavefront reconstruction simulation speed by a factor of 20. As an example for science applications, we study the optimal optical gain, a measurement of pyramid wavefront sensors sensitivity, for different guide object sizes.
10.26698/AO4ELT5.0049- Proceeding PDF