1# npcols MPI grid, 0 leaves MPI to find the best grid. 2# Note that the total number of processors must be divisible per npcols 30 4# Use MPI-RMA 5T 6# operation 7dbcsr_multiply 8# matrix sizes (M, N, K) 91000 101000 111000 12# sparsity (A, B, C) 130.9d0 140.9d0 150.9d0 16# transposes 17N 18N 19# symmetries 20N 21N 22N 23# data type 243 25# alpha (real, imag) 261.0d0 270.0d0 28# beta (real, imag) 291.0d0 300.0d0 31# limits (0 means full size) 32# row 330 340 35# col 360 370 38# k 390 400 41# retain sparsity (T/F) 42F 43# number of repetitions 4410 45# number of different blocks to read (m, n, k) 461 471 481 49# the m blocks (multiplicity, block size, ...) 501 515 52# the n blocks (multiplicity, block size, ...) 531 545 55# the k blocks (multiplicity, block size, ...) 561 575 58# checksum (check, threshold, references) 59T 600.1E-10 610.996149460464712E+07 620.299382963262912E+08 63