D endritic alloy microstructures are formed during a wide range of solidification processes from casting to welding. These microstructures result from a morphological instability of the solid-liquid interface that produces dendrites, which are highly hierarchical branched patterns with primary-, secondary-and higher-order branches. As alloy impurities segregate in the interdendritic liquid during solidification, the spatially inhomogeneous distribution of impurities in the completely solidified alloy is a direct footprint of the dendritic network that formed and coarsened during the solidification process. It also determines the formation and distribution of secondary phases, and thus has a profound influence on the properties of a wide range of technologically important structural materials, from light-weight aluminium alloys used in the automotive industry to nickel-based superalloys used for turbine blades. The study of dendritic growth 1,2 has also been of long-standing fundamental interest because of the ubiquity of branched structures exhibited by diverse interfacial pattern formation systems [3][4][5] .Major theoretical and computational advances over the past two decades have improved our fundamental understanding of dendrite growth, as well as new capabilities to simulate and predict dendritic microstructures on experimentally relevant length and timescales 6 and to elucidate new pattern formation mechanisms 7,8 that enlarge the scope of our understanding of these structures. The commonly accepted microscopic solvability theory of steadystate dendrite growth 9-12 , which builds on the earlier diffusive transport theory of Ivantsov 13 , has led to the understanding that crystalline anisotropy is a crucial parameter that uniquely determines the growth rate and tip radius of dendrites, which is the basic scaling length for the entire dendritic network. Predictions of this theory have been largely validated by phase-field simulations of dendritic evolution over the past few years for both small [14][15][16][17][18][19] and large 20 growth rate. Moreover, molecular dynamics (MD) simulation methods [21][22][23][24][25][26][27][28][29][30] as well as experimental techniques 31,32 have recently been developed to accurately compute anisotropic interfacial properties that control dendritic evolution.Despite this progress, dendrite growth theory remains limited to predicting the steady-state characteristics of dendrites growing along simple crystallographic directions, such as the 100 directions that correspond to the main crystal axes for materials with cubic symmetry, or the six directions in the basal plane of