Execution time of the four processes in the ConKer algorithm, as applied to subsamples of the SDSS DR12 CMASS NGC galaxies (150° < α_g < 210°, 0° < δ_g < 60°). Each point represents a calculation of the 2pcf (l = 0,2,4), 3pcf (l = 0–5), and diagonal npcf up to n = 5. Top: dependence on the number of combined data and random objects, using a grid resolution g_S = 8 h⁻¹ Mpc. Bottom: dependence on the grid resolution g_S for 12.5M objects. The points represent the measured CPU time. The dashed lines are the results of the fit to a power law, with the scaling given in the figure. The solid line (bottom) is the time of the convolution step fit to .