mycpp Code Generation

Measure the speedup from mycpp, and the resource usage.

Source code: oil/mycpp/examples

User Time (milliseconds)

Lower ratios are better.

example name C++ Python C++ : Python
classes 0 23 0.000
fib_iter 7 889 0.008
modules 2 182 0.009
loops 4 300 0.012
fib_recursive 11 864 0.012
asdl_generated 11 378 0.028
containers 4 122 0.031
parse 26 770 0.033
scoped_resource 48 1,023 0.047
tuple_return_value 20 198 0.101
files 7 72 0.103
length 45 212 0.210
cartesian 82 332 0.247
escape 107 345 0.309
cgi 262 513 0.510
varargs 12 20 0.616
control_flow 206 104 1.979

Max Resident Set Size (MB)

Lower ratios are better. We use MB (powers of 10), not MiB (powers of 2).

example name C++ Python C++ : Python
classes 4.5 10.7 0.42
parse 3.8 7.6 0.50
scoped_resource 3.5 7.1 0.50
cartesian 3.5 6.9 0.51
cgi 3.5 6.9 0.51
escape 3.7 7.1 0.52
asdl_generated 3.7 6.9 0.53
fib_iter 3.7 6.9 0.53
length 3.7 6.9 0.53
loops 3.8 7.1 0.54
fib_recursive 3.8 6.9 0.55
files 3.8 6.9 0.55
modules 3.8 6.9 0.55
tuple_return_value 3.8 6.9 0.55
control_flow 3.9 6.9 0.57
containers 28.7 48.0 0.60
varargs 5.5 7.1 0.78

System Time (milliseconds)

Lower ratios are better.

example name C++ Python C++ : Python
asdl_generated 0 4 0.000
control_flow 0 8 0.000
escape 0 8 0.000
fib_iter 0 8 0.000
fib_recursive 0 16 0.000
files 0 4 0.000
loops 0 8 0.000
modules 0 8 0.000
scoped_resource 0 12 0.000
tuple_return_value 0 8 0.000
classes 3 8 0.419
containers 11 12 0.969
cgi 4 4 0.989
varargs 53 52 1.027
cartesian 8 4 1.949
parse 4 0 inf
length 0 0 NA

raw benchmark files