Historical Results
Based on (2) saved test runs, The test labeled "clean code" was (on average) the faster by 0.0141 seconds, (1.299% faster)
The clean code test took 1.0684 seconds.
The notices test took 1.0824 seconds.
Nitty-Gritty
Each test case ran 20 random code order iterations consisting of 275,691 loops for a total of 5,513,820 runs.
- Line execution difference (0.000003) milliseconds.
- Avg difference (0.000) milliseconds per 275,691 loops.
- Total difference 14.06 milliseconds for 5,513,820 loops
The iteration variablity for Code 1 was (0.0000) milliseconds and Code 2 was (0.0000) milliseconds. The lower and the closer together there values are the more accurate the results are.
Code
The first test, "notices", was:
for ($i = 0; $i < 100; $i++) @$GLOBALS[dummy]++;
The second test, "clean code", was:
for ($i = 0; $i < 100; $i++) $GLOBALS['dummy']++;