Knowledge

Livermore loops

Source 📝

200:
F. H. McMahon. Livermore fortran kernels: A computer test of numerical performance range. Technical Report UCRL-53745, Lawrence Livermore National Laboratory, Livermore, CA, December 1986. NTIS report #
43: 258: 218: 253: 171: 154: 176: 125: 115: 78: 166: 120: 31: 213:
Xingfu Wu. Performance Evaluation, Prediction and Visualization of Parallel Systems. Springer, 1999.
133: 147: 81: 35: 214: 108: 102: 201: 247: 92: 86: 71: 47: 39: 238: 233: 67:
Each loop carries out a different mathematical kernel . Those kernels are:
55:
Livermore fortran kernels: A computer test of numerical performance range
61: 50:, some of which can be vectorized, and some of which cannot. 64:, but have since been ported to many programming languages. 38:. It was created by Francis H. McMahon from scientific 60:The Livermore loops were originally written in 8: 16:Type of benchmark for parallel computers 193: 53:The benchmark was published in 1986 in 44:Lawrence Livermore National Laboratory 7: 181:2-D implicit hydrodynamics fragment 162:general linear recurrence equations 159:2-D explicit hydrodynamics fragment 98:tridiagonal linear systems solution 184:location of a first array minimum. 14: 234:Livermore Loops, Fortran version 1: 275: 239:Livermore Loops, C version 24:Livermore Fortran kernels 259:Supercomputer benchmarks 172:matrix-matrix transport 155:conditional computation 177:Planckian distribution 141:1-D particle in a cell 138:2-D particle in a cell 114:alternating direction 126:difference predictors 121:integrate predictors 116:implicit integration 46:. It consists of 24 42:run on computers at 167:ordinates transport 22:(also known as the 254:Parallel computing 82:conjugate gradient 36:parallel computers 109:equation of state 103:linear recurrence 266: 222: 211: 205: 198: 134:first difference 274: 273: 269: 268: 267: 265: 264: 263: 244: 243: 230: 225: 212: 208: 199: 195: 191: 20:Livermore loops 17: 12: 11: 5: 272: 270: 262: 261: 256: 246: 245: 242: 241: 236: 229: 228:External links 226: 224: 223: 206: 192: 190: 187: 186: 185: 182: 179: 174: 169: 163: 160: 157: 151: 145: 144:casual Fortran 142: 139: 136: 131: 128: 123: 118: 112: 106: 99: 96: 93:linear systems 89: 84: 75: 15: 13: 10: 9: 6: 4: 3: 2: 271: 260: 257: 255: 252: 251: 249: 240: 237: 235: 232: 231: 227: 220: 219:0-7923-8462-8 216: 210: 207: 203: 197: 194: 188: 183: 180: 178: 175: 173: 170: 168: 164: 161: 158: 156: 152: 149: 146: 143: 140: 137: 135: 132: 129: 127: 124: 122: 119: 117: 113: 110: 107: 104: 100: 97: 94: 90: 88: 87:inner product 85: 83: 80: 76: 73: 72:hydrodynamics 70: 69: 68: 65: 63: 58: 56: 51: 49: 45: 41: 37: 33: 29: 25: 21: 209: 196: 66: 59: 54: 52: 27: 23: 19: 18: 221:. Page 144. 148:Monte Carlo 77:incomplete 40:source code 248:Categories 202:DE87009360 189:References 165:discrete 153:implicit 130:first sum 105:equations 32:benchmark 111:fragment 101:general 95:solution 79:Cholesky 74:fragment 48:do loops 91:banded 62:Fortran 30:) is a 217:  150:search 215:ISBN 34:for 28:LFK 26:or 250:: 57:. 204:.

Index

benchmark
parallel computers
source code
Lawrence Livermore National Laboratory
do loops
Fortran
hydrodynamics
Cholesky
conjugate gradient
inner product
linear systems
linear recurrence
equation of state
implicit integration
integrate predictors
difference predictors
first difference
Monte Carlo
conditional computation
ordinates transport
matrix-matrix transport
Planckian distribution
DE87009360
ISBN
0-7923-8462-8
Livermore Loops, Fortran version
Livermore Loops, C version
Categories
Parallel computing
Supercomputer benchmarks

Text is available under the Creative Commons Attribution-ShareAlike License. Additional terms may apply.