Fortran Execution Time Benchmarks: Linux64 on AMD Phenom II

Download Trial Software

Absoft(AP)
15.0.0
Absoft
15.0.0
G95
0.93
GFortran
4.9
Intel(AP)
15.0
Intel
15.0
Lahey
8.10b
PGI
14.9
Oracle
8.6

AC
7.49 7.54 14.66 8.31 9.57 9.48 11.56 8.70 29.20

AERMOD
17.35 17.02 38.81 27.90 16.90 17.00 18.26 16.99 15.46

AIR
4.46 5.21 11.56 6.68 2.13 4.57 4.82 6.62 4.95

CAPACITA
37.23 35.35 56.79 46.84 37.28 37.72 42.78 30.46 51.19

CHANNEL2
154.85 159.54 468.22 166.02 109.99 163.15 237.35 157.81 158.04

DODUC
26.67 26.32 35.31 26.57 22.82 22.47 27.06 25.49 22.88

FATIGUE2
85.04 80.92 559.28 73.14 63.89 64.97 162.95 116.01 114.38

GAS_DYN2
131.63 110.36 449.04 155.42 100.58 100.36 167.31 156.32 120.97

INDUCT2
57.61 97.22 281.32 89.65 43.10 109.35 154.40 196.48 190.58

LINPK
9.89 9.83 11.39 11.24 10.23 10.37 11.24 10.92 10.89

MDBX
12.80 12.96 13.26 11.93 10.70 10.73 12.40 13.03 11.99

MP_PROP_DESIGN
71.29 135.02 711.21 161.44 34.59 107.42 202.11 125.47 209.06

NF
13.83 12.90 24.70 12.35 12.12 12.41 18.41 12.96 13.09

PROTEIN
31.25 31.11 44.71 31.68 29.73 29.83 48.20 36.27 31.64

RNFLOW
20.97 17.36 38.76 19.58 18.10 18.58 25.08 23.01 26.81

TEST_FPU2
90.54 103.57 174.64 98.15 83.22 88.63 109.93 92.79 105.66

TFFT2
131.30 132.66 137.96 89.78 135.68 131.95 128.27 134.34 128.99
Geometric Mean 32.70 34.54 73.08 37.09 27.23 33.13 44.47 38.81 42.33
Compiler Switches
Absoft(AP) af90 -m64 -O5 -speed_math=10 -march=barcelona -xINTEGER
(These settings enable auto-parallelization, see below)
Absoft af90 -m64 -O4 -speed_math=10 -march=barcelona -xINTEGER
g95 g95 -march=opteron -ffast-math -funroll-loops -O3
gfortran gfortran -ffast-math -funroll-loops -O3
Intel(AP) ifort -O3 -fast -parallel -ipo (These settings enable auto-parallelization, see below)
Intel ifort -O3 -fast -ipo
Lahey lf95 –fast -static -x -
PGI pgf90 -V -fast -Mipa=fast,inline -Msmartalloc -Mfprelaxed -Mstack_arrays
Oracle sunf95 -fast -xtarget=native
Notes
All figures are Execution Times in Seconds – measured on a machine with an AMD Phenom II X4 955 processor (3.2 GHz), with 4GBytes memory, running CentOS 5.5. Each figure is the average over at least 10 runs (many more for some). Measurement error is typically <1%.  Green cells highlight figures within 10% of the fastest.  Red cells indicate figures which are more than 150% of the fastest.So far as possible, we have used the compiler switches which give the best overall results.  We have not attempted to tune individual benchmarks, and, in particular cases, different switch settings may give better results.The settings used for the Intel and Absoft compilers enable autoparallelization.  Autoparallelization settings are not used on any other compilers because we found that they produced no significant performance benefits on this benchmark set.

Thanks are due to Jos Bergervoet for permission to use his CAPACITA benchmark, to Quetzal Associates for permission to use their CHANNEL, FATIGUE, GAS_DYN, INDUCT, PROTEIN and RNFLOW benchmarks, to David Frank for his TEST_FPU benchmark, to Anthony Falzone for the use of MP_PROP_DESIGN, and to Ted Addison of McVehil-Monnett Associates for permission to use AERMOD, an air quality model used by the US Environmental Protection Agency.

All the benchmarks have been modified slightly to fit into our benchmarking harness.

The NF benchmark uses  “nested factorization”, a little known but very effective iterative linear solver for huge finite difference matrices.  A paper describing nested factorization, and comparing it to other methods is available here.

 

Other benchmark sets:

AMD/Windows

Intel/Windows (Sandy Bridge)

Intel/Linux

 

 

Download Polyhedron Benchmarks

Download Trial Software

 

Return to Fortran Compiler Comparisons homepage