PerlBench CPU AMD Opteron 6128

Performed by: VX Solutions s.r.o.

Date: 11.03.2011

  lab1 lab2 lab3 lab4 lab5 lab6 lab7 lab8 lab9 lab10
arith/mixed

96

97

96

100

100

100

96

97

91

94

arith/trig

100

97

96

100

100

100

100

100

100

99

array/copy

100

99

100

100

100

102

90

91

90

92

array/foreach

92

95

92

100

100

100

91

91

90

88

array/index

98

100

100

100

100

100

85

90

89

85

array/pop

100

100

99

100

100

100

86

86

88

87

array/shift

100

100

97

100

100

100

86

85

87

85

array/sort

98

100

98

100

100

100

80

80

80

80

array/sort-num

98

97

96

100

100

100

84

84

85

84

call/0arg

100

98

100

100

100

102

90

90

91

90

call/1arg

98

97

100

100

100

100

88

88

89

87

call/2arg

93

93

95

97

100

99

84

84

86

83

call/9arg

95

97

98

100

100

88

88

87

87

87

call/empty

95

91

95

100

100

100

86

89

89

89

call/fib

93

92

95

100

100

100

85

83

85

85

call/method

100

100

100

105

100

100

85

82

84

82

call/wantarray

98

95

99

100

100

100

89

87

89

88

hash/copy

100

100

100

101

100

105

93

93

94

94

hash/each

101

100

103

107

100

111

100

98

100

99

hash/foreach

99

99

100

101

100

102

83

85

84

85

hash/foreach-sort

96

97

94

100

100

100

78

78

78

78

hash/get

97

92

96

100

100

100

84

86

84

86

hash/set

96

94

95

100

100

100

80

84

81

83

loop/for-c

97

99

100

100

100

102

93

93

95

94

loop/for-range

95

94

94

100

100

100

93

95

91

92

loop/for-range-const

94

94

92

100

100

100

93

93

90

91

loop/getline

95

97

100

96

100

100

92

91

90

92

loop/while

95

97

95

100

100

101

89

90

90

90

loop/while-my

93

95

96

100

100

100

93

88

87

97

re/const

94

100

98

94

100

100

90

93

91

92

re/w

106

100

103

96

100

103

99

100

100

100

startup/fewmod

100

100

100

100

100

100

97

96

96

97

startup/lotsofsub

100

100

100

100

100

101

91

91

90

90

startup/noprog

100

100

100

100

100

100

100

100

100

100

string/base64

95

86

94

100

100

101

86

85

86

85

string/htmlparser

99

96

98

100

100

103

86

86

86

85

string/index-const

93

99

98

100

100

100

92

92

90

90

string/index-var

128

100

111

91

100

100

119

119

117

119

string/ipol

100

100

100

100

100

100

95

96

97

94

string/tr

95

95

100

100

100

100

93

94

93

94

AVERAGE 98 97 98 100 100 100 90 91 90 90
Results are in percent relative to lab5. Each test measured 5 times, values adjusted for 95% confidence.

Configuration summary

Test ran on a freebsd machine that reports its uname as "FreeBSD 8.2-RELEASE #0 r218742M: Mon Feb 21 11:49:51 CET 2011 amd64".

Lab Perl version Compiler Optimization
lab1 5.12.3 4.2.1 20070719 [FreeBSD] -O2 -pipe -fno-strict-aliasing
lab2 5.12.3 4.2.1 20070719 [FreeBSD] -O2 -pipe -march=nocona -fno-strict-aliasing
lab3 5.12.3 4.2.1 20070719 [FreeBSD] -O2 -pipe -march=opteron-sse3 -fno-strict-aliasing
lab4 5.12.3 4.5.3 20110106 (prerelease) -O2 -pipe -Wl,-rpath=/usr/local/lib/gcc45 -fno-strict-aliasing
lab5 5.12.3 4.5.3 20110106 (prerelease) -O2 -pipe -march=nocona -Wl,-rpath=/usr/local/lib/gcc45 -fno-strict-aliasing
lab6 5.12.3 4.5.3 20110106 (prerelease) -O2 -pipe -march=barcelona -Wl,-rpath=/usr/local/lib/gcc45 -fno-strict-aliasing
lab7 5.12.3 4.2.1 Compatible Clang Compiler -O2 -pipe -fno-strict-aliasing
lab8 5.12.3 4.2.1 Compatible Clang Compiler -O2 -pipe -march=nocona -fno-strict-aliasing
lab9 5.12.3 4.2.1 Compatible Clang Compiler -O2 -pipe -march=opteron-sse3 -fno-strict-aliasing
lab10 5.12.3 4.2.1 Compatible Clang Compiler -O2 -pipe -march=barcelona -fno-strict-aliasing
Clang revision: 127334

Back to summary


blog.vx.sk
Martin Matuška <>