1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
|
# npcols MPI grid, 0 leaves MPI to find the best grid.
# Note that the total number of processors must be divisible per npcols
0
# Use MPI-RMA
F
# operation
dbcsr_multiply
# matrix sizes (M, N, K)
50
50
50
# sparsity (A, B, C)
0.0d0
0.0d0
0.0d0
# transposes
N
N
# symmetries
N
N
N
# data type
3
# alpha (real, imag)
1.0d0
0.0d0
# beta (real, imag)
1.0d0
0.0d0
# limits (0 means full size)
# row
0
0
# col
0
0
# k
0
0
# retain sparsity (T/F)
F
# number of repetitions
10
# number of different blocks to read (m, n, k)
1
1
1
# the m blocks (multiplicity, block size, ...)
1
50
# the n blocks (multiplicity, block size, ...)
1
50
# the k blocks (multiplicity, block size, ...)
1
50
# checksum (check, threshold, references)
T
0.1E-10
0.418186760034529E+06
0.190157258297048E+06
|
