Dr Patrick H. Worley Oak Ridge National Laboratory System Description ------------------ 11/17/95 XPS150 if77/Paragon Paragon Version R5.2 if77 flags: -O4 -Mnodepchk -Knoieee Paragon Software Release R1.3.3 mpich 1.0.11 (arch=paragon device=nx) Timing Tables ------------- make MACH=paragon COMM=mpich WORKSPACE=6700000 optimization level 0 optimization level 1 T42L16 1x1 1954.9 -sz 1x1 +1 2x1 1148.1 -sz 1x2 +1 1x2 978.54 -sz 2x1 +1 2x2 577.73 -sz 2x2 +1 1x4 495.70 -sz 4x1 +1 4x2 293.77 -sz 2x4 +1 2x4 290.70 -sz 4x2 +1 4x4 145.98 -sz 4x4 +1 4x4 145.00 -sz 4x4 +1 8x4 73.86 -sz 4x8 +1 8x4 73.50 -sz 4x8 +1 8x8 39.43 -sz 8x8 +1 8x8 38.38 -sz 8x8 +1 16x8 22.44 -sz 8x16 +1 16x8 21.49 -sz 8x16 +1 16x16 15.62 -sz 16x16 +1 16x16 13.97 -sz 16x16 +1 T170L32 8x4 4405.5 -sz 4x8 +1 8x8 2226.3 -sz 8x8 +1 16x8 1129.0 -sz 8x16 +1 16x16 605.02 -sz 16x16 +1 16x16 600.76 -sz 16x16 +1 32x16 342.93 -sz 16x32 +1 32x16 342.67 -sz 16x32 +1 32x32 425.27 -sz 16x64 +1 32x32 263.85 -sz 16x64 +0 32x32 299.65 -sz 16x64 +0 (*) T340L64 16x16 7927.7 -sz 16x16 +1 32x16 4336.3 -sz 16x32 +1 32x16 4192.1 -sz 16x32 +1 32x32 4310.6 -sz 16x64 +1 64x16 2747.7 -sz 16x64 +1 32x32 3289.5 -sz 16x64 +0 (*) Notes: 1) Mapped logical processor mesh to physcial processor mesh for all processor numbers up to 512 (see -sz specifications). For 1024, can map 32x32 logical mesh directly onto 16x64 physical processor mesh. Results are provided both for the default mapping and for a specific nearest neighbor "preserving" map for level 0 optimization. Level 1 optimization for 1024 processors also uses the improved mapping. 2) Missing optimization level 1 values denote problem sizes in which options better than the default have not yet been identified. Raw Timing Output Data ---------------------- small problem; level 0 optimization 0.19549E+04 0.90064E+01 0.90944E+01 0.00000E+00 216 ( 42, 42, 42) ( 128, 64, 16) (2,T,F,F) ( 1, 1) 1 (1,10,10, 2, 2,6,6) (0, 0, 0, 1, 2,6,6) 0 0 4691369 0.11481E+04 0.52764E+01 0.53318E+01 0.00000E+00 216 ( 42, 42, 42) ( 128, 64, 16) (2,T,F,F) ( 2, 1) 1 (1,10,10, 2, 2,6,6) (0, 0, 0, 1, 2,6,6) 0 0 2412937 0.57773E+03 0.26452E+01 0.26847E+01 0.00000E+00 216 ( 42, 42, 42) ( 128, 64, 16) (2,T,F,F) ( 2, 2) 1 (1,10,10, 2, 2,6,6) (0, 0, 0, 1, 2,6,6) 0 0 1206801 0.29377E+03 0.13459E+01 0.13684E+01 0.00000E+00 216 ( 42, 42, 42) ( 128, 64, 16) (2,T,F,F) ( 4, 2) 1 (1,10,10, 2, 2,6,6) (0, 0, 0, 1, 2,6,6) 0 0 637193 0.14598E+03 0.66934E+00 0.67870E+00 0.00000E+00 216 ( 42, 42, 42) ( 128, 64, 16) (2,T,F,F) ( 4, 4) 1 (1,10,10, 2, 2,6,6) (0, 0, 0, 1, 2,6,6) 0 0 318969 0.73856E+02 0.33810E+00 0.34372E+00 0.00000E+00 216 ( 42, 42, 42) ( 128, 64, 16) (2,T,F,F) ( 8, 4) 1 (1,10,10, 2, 2,6,6) (0, 0, 0, 1, 2,6,6) 0 0 176549 0.39432E+02 0.18044E+00 0.18440E+00 0.00000E+00 216 ( 42, 42, 42) ( 128, 64, 16) (2,T,F,F) ( 8, 8) 1 (1,10,10, 2, 2,6,6) (0, 0, 0, 1, 2,6,6) 0 0 88629 0.22435E+02 0.10203E+00 0.10671E+00 0.00000E+00 216 ( 42, 42, 42) ( 128, 64, 16) (2,T,F,F) ( 16, 8) 1 (1,10,10, 2, 2,6,6) (0, 0, 0, 1, 2,6,6) 0 0 53015 0.15626E+02 0.70585E-01 0.74459E-01 0.00000E+00 216 ( 42, 42, 42) ( 128, 64, 16) (2,T,F,F) ( 16, 16) 1 (1,10,10, 2, 2,6,6) (0, 0, 0, 1, 2,6,6) 0 0 27277 medium problem; level 0 optimization 0.44055E+04 0.20242E+02 0.20394E+02 0.00000E+00 216 (170,170,170) ( 512, 256, 32) (2,T,F,F) ( 8, 4) -1 (1,10,10, 2, 2,6,6) (0, 0, 0, 1, 2,6,6) 0 0 5608617 0.22263E+04 0.10224E+02 0.10306E+02 0.00000E+00 216 (170,170,170) ( 512, 256, 32) (2,T,F,F) ( 8, 8) 1 (1,10,10, 2, 2,6,6) (0, 0, 0, 1, 2,6,6) 0 0 2805673 0.11290E+04 0.51839E+01 0.52276E+01 0.00000E+00 216 (170,170,170) ( 512, 256, 32) (2,T,F,F) ( 16, 8) 1 (1,10,10, 2, 2,6,6) (0, 0, 0, 1, 2,6,6) 0 0 1679629 0.60502E+03 0.27753E+01 0.28016E+01 0.00000E+00 216 (170,170,170) ( 512, 256, 32) (2,T,F,F) ( 16, 16) 1 (1,10,10, 2, 2,6,6) (0, 0, 0, 1, 2,6,6) 0 0 841161 0.34293E+03 0.15728E+01 0.15889E+01 0.00000E+00 216 (170,170,170) ( 512, 256, 32) (2,T,F,F) ( 32, 16) 1 (1,10,10, 2, 2,6,6) (0, 0, 0, 1, 2,6,6) 0 0 573633 0.42527E+03 0.18638E+01 0.20784E+01 0.00000E+00 216 (170,170,170) ( 512, 256, 32) (2,T,F,F) ( 32, 32) 1 (1,10,10, 2, 2,6,6) (0, 0, 0, 1, 2,6,6) 0 0 345993 0.29965E+03 0.13543E+01 0.14112E+01 0.00000E+00 216 (170,170,170) ( 512, 256, 32) (2,T,F,F) ( 32, 32) 0 (1,10,10, 2, 2,6,6) (0, 0, 0, 1, 2,6,6) 0 0 345993 large problem; level 0 optimization 0.79277E+04 0.35815E+02 0.37261E+02 0.00000E+00 216 (340,340,340) (1024, 512, 64) (2,T,F,F) ( 16, 16) 1 (1,10,10, 2, 2,6,6) (0, 0, 0, 1, 2,6,6) 0 0 6510845 0.43363E+04 0.19345E+02 0.20656E+02 0.00000E+00 216 (340,340,340) (1024, 512, 64) (2,T,F,F) ( 32, 16) 1 (1,10,10, 2, 2,6,6) (0, 0, 0, 1, 2,6,6) 0 0 4271209 0.43106E+04 0.18940E+02 0.21041E+02 0.00000E+00 216 (340,340,340) (1024, 512, 64) (2,T,F,F) ( 32, 32) 1 (1,10,10, 2, 2,6,6) (0, 0, 0, 1, 2,6,6) 0 0 2209265 0.32895E+04 0.14923E+02 0.15548E+02 0.00000E+00 216 (340,340,340) (1024, 512, 64) (2,T,F,F) ( 32, 32) 0 (1,10,10, 2, 2,6,6) (0, 0, 0, 1, 2,6,6) 0 0 2209265 small problem; level 1 optimization 0.97854E+03 0.44950E+01 0.45527E+01 0.00000E+00 216 ( 42, 42, 42) ( 128, 64, 16) (2,T,F,F) ( 1, 2) 1 (1,14,14, 2, 2,3,3) (0, 2, 2, 2, 2,3,3) 0 0 2391425 0.49570E+03 0.22741E+01 0.23044E+01 0.00000E+00 216 ( 42, 42, 42) ( 128, 64, 16) (2,T,F,F) ( 1, 4) 1 (1,14,14, 2, 2,3,3) (0, 2, 2, 4, 5,3,3) 0 0 1241745 0.29070E+03 0.13334E+01 0.13555E+01 0.00000E+00 216 ( 42, 42, 42) ( 128, 64, 16) (2,T,F,F) ( 2, 4) 1 (1,14,14, 2, 2,3,3) (0, 2, 2, 4, 5,3,3) 0 0 637937 0.14500E+03 0.66348E+00 0.67443E+00 0.00000E+00 216 ( 42, 42, 42) ( 128, 64, 16) (2,T,F,F) ( 4, 4) 1 (1,14,14, 2, 2,3,3) (0, 2, 2, 4, 5,3,3) 0 0 336033 0.73502E+02 0.33567E+00 0.34176E+00 0.00000E+00 216 ( 42, 42, 42) ( 128, 64, 16) (2,T,F,F) ( 8, 4) -1 (1,14,14, 2, 2,3,3) (0, 2, 2, 4, 5,3,3) 0 0 185081 0.38378E+02 0.17487E+00 0.17971E+00 0.00000E+00 216 ( 42, 42, 42) ( 128, 64, 16) (2,T,F,F) ( 8, 8) 1 (1,14,14, 2, 2,3,3) (0, 2, 2, 8, 9,3,3) 0 0 98625 0.21494E+02 0.97340E-01 0.10207E+00 0.00000E+00 216 ( 42, 42, 42) ( 128, 64, 16) (2,T,F,F) ( 16, 8) -1 (1,14,14, 2, 2,3,3) (0, 2, 2, 8, 9,3,3) 0 0 58013 0.13967E+02 0.62490E-01 0.67346E-01 0.00000E+00 216 ( 42, 42, 42) ( 128, 64, 16) (2,T,F,F) ( 16, 16) 1 (1,14,14, 2, 2,3,3) (0, 2, 2, 16, 17,3,3) 0 0 32677 medium problem; level 1 optimization 0.60076E+03 0.27310E+01 0.28225E+01 0.00000E+00 216 (170,170,170) ( 512, 256, 32) (2,T,F,F) ( 16, 16) 1 (1,14,14, 2, 2,3,3) (0, 2, 2, 16, 17,3,3) 0 0 1006761 0.34267E+03 0.15373E+01 0.16163E+01 0.00000E+00 216 (170,170,170) ( 512, 256, 32) (2,T,F,F) ( 32, 16) 1 (1,14,14, 2, 2,3,3) (0, 2, 2, 16, 17,3,3) 0 0 656433 0.26385E+03 0.11803E+01 0.12504E+01 0.00000E+00 216 (170,170,170) ( 512, 256, 32) (2,T,F,F) ( 32, 32) 0 (1,12,12, 2, 2,3,3) (0, 2, 2, 32, 33,3,3) 0 0 431553 large problem; level 1 optimization 0.41921E+04 0.18326E+02 0.20423E+02 0.00000E+00 216 (340,340,340) (1024, 512, 64) (2,T,F,F) ( 32, 16) 1 (1,14,14, 2, 2,3,3) (0, 2, 2, 16, 17,3,3) 0 0 4927309 0.28906E+04 0.11922E+02 0.14970E+02 0.00000E+00 216 (340,340,340) (1024, 512, 64) (2,T,F,F) ( 32, 32) 0 (1,14,14, 2, 2,3,3) (0, 2, 2, 32, 33,3,3) 0 0 2887421 0.27477E+04 0.12569E+02 0.12757E+02 0.00000E+00 216 (340,340,340) (1024, 512, 64) (2,T,F,F) ( 64, 16) 1 (1,10,10, 2, 2,6,6) (0, 0, 0, 1, 2,6,6) 0 0 3205999