Speaker
Description
X-ray Photon Correlation Spectroscopy (XPCS) inverts the dynamic behavior of materials at the nanoscale by analyzing the "photon intensity fluctuations" in synchrotron X-ray scattering signals. The multitau algorithm (multi-tau algorithm) calculates lag times at different time levels through a "hierarchical grouping" strategy, which can cover a wide time range while controlling computational complexity. When computing the one-time correlation function and two-time correlation function, the input data are ultra-large-scale datasets (n>10000), leading to low computational efficiency that often requires several hours or even days to complete. In this study, the acceleration of correlation function calculations is realized based on CPU parallel computing and GPU parallel computing, and the methods of data partitioning and computing sequence allocation are described. Experimental results show that CPU parallel computing achieves an average 10-fold speedup, and GPU parallel computing achieves an average 80-fold speedup. Meanwhile, an XPCS data processing platform is presented, which supports interactive data processing.
