Intensity of short-distance contacts is 10-1000 times larger than that of long-distance, resulting in the intense signals on diagonal but sparse away on Hi-C maps. Here we developed NorDis, which accounted for distance effects and balanced the short-distance and long-distance signals. In NorDis, scaled genomic distances (SGD) are used to calculate the average contact number. SGD includes more points with larger distance, so that it would not be affected by outliers. After NorDis, chromosome-level architecture global patterns showed up, revealing how chromosomes are organized in nuclei. We developed a clustering methods using sparse dimension reduction to discover the widely-distributed patterns. At results, we found 6 intra-chromosome and 5 inter-chromosome global patterns in 1,032 species. The other high-order structure we found, local texture, however, shows appearance in complex species but not in simple species. By developing an algorithm measuring the strength of local textures, we found the strength correlates with organism complexity. Furthermore, we found that local textures independently evolved in plants, invertebrates and vertebrates. With closer looking at Actinopteri, recurrent evolution also shows in each order. Noteworthy, the local texture strength correlates with cell type numbers and organism lengths, which are always regarded as the measurements of organism complexity.