Microbenchmarking in Java
// let's benchmark the creation of 1,000,000 dummy objects
Dummy.createNDummies(1000000);
// dummy
public static void createNDummies(int n) {
for (int i = 0; i < n; i++) {
Dummy dummy = new Dummy();
dummy.doNothing();
}
}Times without manual warmup
|
Only Interpreter |
C1 |
C2 |
C1 & C2 |
| Time (ms) |
139 |
24 |
6 |
13 |
|
Only Interpreter |
C1 |
C2 |
C1 & C2 |
| Warmup Time (ms) |
106 |
24 |
5 |
12 |
| Execution Time (ms) |
101 |
14 |
0 |
3 |
Interpreter always runs so JIT compiler's time also includes interpreter's time.
Benchmark Mode Cnt Score Error Units
_10MillionsOfDummy_OnlyInterpreter_JMH.createNDummies thrpt 16.822 ops/s
_10MillionsOfDummy_OnlyInterpreter_JMH.createNDummies avgt 0.059 s/op
_10MillionsOfDummy_OnlyInterpreter_JMH.createNDummies sample 170 0.059 ± 0.001 s/op
_10MillionsOfDummy_OnlyInterpreter_JMH.createNDummies:p0.00 sample 0.056 s/op
_10MillionsOfDummy_OnlyInterpreter_JMH.createNDummies:p0.50 sample 0.058 s/op
_10MillionsOfDummy_OnlyInterpreter_JMH.createNDummies:p0.90 sample 0.063 s/op
_10MillionsOfDummy_OnlyInterpreter_JMH.createNDummies:p0.95 sample 0.067 s/op
_10MillionsOfDummy_OnlyInterpreter_JMH.createNDummies:p0.99 sample 0.081 s/op
_10MillionsOfDummy_OnlyInterpreter_JMH.createNDummies:p0.999 sample 0.084 s/op
_10MillionsOfDummy_OnlyInterpreter_JMH.createNDummies:p0.9999 sample 0.084 s/op
_10MillionsOfDummy_OnlyInterpreter_JMH.createNDummies:p1.00 sample 0.084 s/op
_10MillionsOfDummy_OnlyInterpreter_JMH.createNDummies ss 0.065 s/op
Benchmark Mode Cnt Score Error Units
_09MillionsOfDummy_C1C2_JMH.createNDummies thrpt 1770525516.073 ops/s
_09MillionsOfDummy_C1C2_JMH.createNDummies avgt ≈ 10⁻⁹ s/op
_09MillionsOfDummy_C1C2_JMH.createNDummies sample 240034 ≈ 10⁻⁸ s/op
_09MillionsOfDummy_C1C2_JMH.createNDummies:p0.00 sample ≈ 10⁻⁸ s/op
_09MillionsOfDummy_C1C2_JMH.createNDummies:p0.50 sample ≈ 10⁻⁸ s/op
_09MillionsOfDummy_C1C2_JMH.createNDummies:p0.90 sample ≈ 10⁻⁷ s/op
_09MillionsOfDummy_C1C2_JMH.createNDummies:p0.95 sample ≈ 10⁻⁷ s/op
_09MillionsOfDummy_C1C2_JMH.createNDummies:p0.99 sample ≈ 10⁻⁷ s/op
_09MillionsOfDummy_C1C2_JMH.createNDummies:p0.999 sample ≈ 10⁻⁷ s/op
_09MillionsOfDummy_C1C2_JMH.createNDummies:p0.9999 sample ≈ 10⁻⁵ s/op
_09MillionsOfDummy_C1C2_JMH.createNDummies:p1.00 sample ≈ 10⁻⁴ s/op
_09MillionsOfDummy_C1C2_JMH.createNDummies ss 0.007 s/op
Times when ignoring the sum computed
Benchmark Mode Cnt Score Error Units
_11Sum_C1C2_Challenge.sum thrpt 1472.511 ops/s
_11Sum_C1C2_Challenge.sum avgt 0.001 s/op
_11Sum_C1C2_Challenge.sum sample 14934 0.001 ± 0.001 s/op
_11Sum_C1C2_Challenge.sum:p0.00 sample 0.001 s/op
_11Sum_C1C2_Challenge.sum:p0.50 sample 0.001 s/op
_11Sum_C1C2_Challenge.sum:p0.90 sample 0.001 s/op
_11Sum_C1C2_Challenge.sum:p0.95 sample 0.001 s/op
_11Sum_C1C2_Challenge.sum:p0.99 sample 0.001 s/op
_11Sum_C1C2_Challenge.sum:p0.999 sample 0.002 s/op
_11Sum_C1C2_Challenge.sum:p0.9999 sample 0.003 s/op
_11Sum_C1C2_Challenge.sum:p1.00 sample 0.003 s/op
_11Sum_C1C2_Challenge.sum ss 0.005 s/op
Times when considering the sum computed
Benchmark Mode Cnt Score Error Units
_12Sum_C1C2_Solution.sum thrpt 791.332 ops/s
_12Sum_C1C2_Solution.sum avgt 0.001 s/op
_12Sum_C1C2_Solution.sum sample 8054 0.001 ± 0.001 s/op
_12Sum_C1C2_Solution.sum:p0.00 sample 0.001 s/op
_12Sum_C1C2_Solution.sum:p0.50 sample 0.001 s/op
_12Sum_C1C2_Solution.sum:p0.90 sample 0.002 s/op
_12Sum_C1C2_Solution.sum:p0.95 sample 0.002 s/op
_12Sum_C1C2_Solution.sum:p0.99 sample 0.002 s/op
_12Sum_C1C2_Solution.sum:p0.999 sample 0.003 s/op
_12Sum_C1C2_Solution.sum:p0.9999 sample 0.004 s/op
_12Sum_C1C2_Solution.sum:p1.00 sample 0.004 s/op
_12Sum_C1C2_Solution.sum ss 0.006 s/op
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent